To synthesize updated evidence on the cost-effectiveness (CE) of interventions to manage diabetes, its complications, and comorbidities.
We conducted a systematic literature review of studies from high-income countries evaluating the CE of diabetes management interventions recommended by the American Diabetes Association (ADA) and published in English between June 2008 and July 2017. We also incorporated studies from a previous CE review from the period 1985–2008. We classified the interventions based on their strength of evidence (strong, supportive, or uncertain) and levels of CE: cost-saving (more health benefit at a lower cost), very cost-effective (≤$25,000 per life year gained [LYG] or quality-adjusted life year [QALY]), cost-effective ($25,001–$50,000 per LYG or QALY), marginally cost-effective ($50,001–$100,000 per LYG or QALY), or not cost-effective (>$100,000 per LYG or QALY). Costs were measured in 2017 U.S. dollars.
Seventy-three new studies met our inclusion criteria. These were combined with 49 studies from the previous review to yield 122 studies over the period 1985–2017. A large majority of the ADA-recommended interventions remain cost-effective. Specifically, we found strong evidence that the following ADA-recommended interventions are cost-saving or very cost-effective: In the cost-saving category are 1) ACE inhibitor (ACEI)/angiotensin receptor blocker (ARB) therapy for intensive hypertension management compared with standard hypertension management, 2) ACEI/ARB therapy to prevent chronic kidney disease and/or end-stage renal disease in people with albuminuria compared with no ACEI/ARB therapy, 3) comprehensive foot care and patient education to prevent and treat foot ulcers among those at moderate/high risk of developing foot ulcers, 4) telemedicine for diabetic retinopathy screening compared with office screening, and 5) bariatric surgery compared with no surgery for individuals with type 2 diabetes (T2D) and obesity (BMI ≥30 kg/m2). In the very cost-effective category are 1) intensive glycemic management (targeting A1C <7%) compared with conventional glycemic management (targeting an A1C level of 8–10%) for individuals with newly diagnosed T2D, 2) multicomponent interventions (involving behavior change/education and pharmacological therapy targeting hyperglycemia, hypertension, dyslipidemia, microalbuminuria, nephropathy/retinopathy, secondary prevention of cardiovascular disease with aspirin) compared with usual care, 3) statin therapy compared with no statin therapy for individuals with T2D and history of cardiovascular disease, 4) diabetes self-management education and support compared with usual care, 5) T2D screening every 3 years starting at age 45 years compared with no screening, 6) integrated, patient-centered care compared with usual care, 7) smoking cessation compared with no smoking cessation, 8) daily aspirin use as primary prevention for cardiovascular complications compared with usual care, 9) self-monitoring of blood glucose three times per day compared with once per day among those using insulin, 10) intensive glycemic management compared with conventional insulin therapy for T2D among adults aged ≥50 years, and 11) collaborative care for depression compared with usual care.
Complementing professional treatment recommendations, our systematic review provides an updated understanding of the potential value of interventions to manage diabetes and its complications and can assist clinicians and payers in prioritizing interventions and health care resources.
Introduction
Diabetes is a serious, common, and costly disease, affecting 34 million Americans (1) and leading to $327 billion (2) in annual health expenditures in 2017. To better manage and lower the burdens of diabetes, the American Diabetes Association (ADA) annually publishes its Standards of Medical Care in Diabetes (SOC) (3), the most comprehensive and up-to-date clinical knowledge regarding diabetes care. Worldwide, the SOC provides clinicians, patients, researchers, and payers with the most current evidence-based screening, diagnostic, prevention, and management recommendations for diabetes. However, the cost-effectiveness (CE) of these strategies—in other words, the value they provide for these investments—varies greatly and should be considered in management or policy decisions.
CE analysis is an analytical framework that weighs the benefits and costs of an intervention by comparing it with standard care or other alternatives to see if the value of the intervention justifies its cost. The CE of different treatment options provides critical information to stakeholders at a variety of levels; this information helps in development of optimal treatment strategies or policies to lower current and future health and economic burdens and help determine use of limited health care resources.
In 2010, Li et al. (4) systematically reviewed all English-language studies published between January 1985 and May 2008 on the CE of diabetes prevention and management interventions recommended by the ADA’s SOC 2008 (5). The authors concluded that a large majority of the interventions recommended by ADA at that time were cost-saving or very cost-effective. Since then, however, many new technologies and medications have become available and have been added to updated iterations of the ADA’s SOC, leading to important changes in the ways diabetes care is delivered, how complications are managed, and the resulting costs. Recently published CE studies on these new technologies and medications can provide evidence on how to prioritize these novel interventions together with older strategies that remain cost-effective.
To provide up-to-date guidance that aligns with the 2019 SOC (the most up-to-date version at the time of data analysis) (6,7), we aggregated all available data published in English regarding the CE of ADA-recommended interventions to identify diabetes or gestational diabetes mellitus, manage diabetes, screen for diabetes complications, and manage complications and comorbidities. We included data from the previous review by Li et al. and systematically added data from the last decade to provide findings that are relevant to contemporary clinical practice. The result allows us to assess economic implications of changes that have occurred in the way diabetes care is delivered and how complications are managed. As a complement to the ADA’s 2019 SOC recommendations, findings from this review could assist clinicians and policy makers in selecting interventions that are not only effective but also deliver value.
Research Design and Methods
Data Sources and Literature Search
We followed the same search strategy that was used in the 2010 review by Li et al. (4). Briefly, we performed a thorough literature search of seven databases (Cumulative Index to Nursing and Allied Health Literature [CINAHL], Cochrane, EMBASE, MEDLINE, PsycINFO, Scopus, and Sociological Abstracts [Soc Abs]) following the Cochrane Collaboration’s protocol (8) covering the period of June 2008 to July 2017. Medical subject headings matching the previous review’s search protocol were selected to create a search strategy. Our search terms were diabetes (indicating the disease of diabetes: “type 1,” “type 2,” “gestational,” “impaired glucose,” and “insulin resistance”), costs (“cost or expenditure,” “health care cost,” “costs or cost analysis”), effectiveness (“benefit” OR “life year” OR “quality-adjusted life years” OR “disability adjusted life years”), cost-effectiveness (indicating CE analysis, such as “cost-effectiveness analysis” OR “cost-utility analysis” OR “economic”). We also screened reference lists of all included articles and publications from leading medical and diabetes-focused journals during the period for additional articles that may have been missed.
Article Screening and Selection
We included studies from populations in high-income countries (based on World Bank classifications [9]) that assessed the economic value associated with diabetes management interventions included in the ADA’s SOC 2019 (7), the most up-to-date version available at the time of our analysis. We included studies of patients with undiagnosed or diagnosed diabetes, including type 1 diabetes (T1D), type 2 diabetes (T2D), and gestational diabetes mellitus.
We included studies that used one of the three major types of economic evaluations: cost-effectiveness analyses, cost-utility analyses, or cost-benefit analyses, with outcomes measured as cost per additional quality-adjusted life year (QALY) gained, cost per additional life year gained (LYG), or cost per additional disability-adjusted life year averted.
We included original research studies published in English and excluded review articles, commentaries or letters, conference abstracts, and dissertations. For this review, we also excluded studies that focused on preventing T2D. Each study was screened for eligibility by two authors (K.R.S., X.Zho., B.P.N., S.J., K.P., and X.Zha. all participated in this stage), with disagreements resolved by group discussion and consensus.
Data Abstraction
For studies from June 2008 through July 2017, we used the same detailed data extraction form used by Li et al. (4) to systematically gather the following information from each included study: publication information (title, first author’s name, publication year), study population, intervention and comparison, study method (within trial versus modeled/simulation), analytical time horizon and discounting, perspective used, costs, health outcome measures, survey instruments for measuring utility, incremental cost-effectiveness ratio (ICER), whether a sensitivity analysis was conducted, and study conclusion(s). For studies from January 1985 through May 2008, we used the previously abstracted data from Li et al.
Quality Assessment of Included Studies
To evaluate the quality of the included CE studies, we selected a widely used quality assessment tool based on The BMJ authors’ guide for economic studies (10), which was used in the 2010 review by Li et al. We considered four quality score items: source of cost-specific data, categories of costs, source of benefit-specific data, and categories of benefits. We classified studies that did not report all four items as “low quality.” Of studies that had all four components, we assessed nine additional items (analytical time horizon, study perspective, description of the CE model, diagram for constructing the decision tree, currency and year of cost, cost discounting, benefit discounting, ICERs, and sensitivity analysis) for further classification. We assigned one point for each item that was reported. We rated each study as “fair” if it scored 3 or less, “good” if it scored 4–6, and “excellent” if it scored 7–9 points. We restricted our analysis to articles rated as “excellent” or “good” quality. Our quality assessment methodology mirrored that used by Li et al. Figure 1 illustrates the complete process for screening articles.
Data Analysis and Synthesis
To synthesize findings, we first grouped articles into four broad categories: 1) screening for undiagnosed diabetes (including T2D and gestational diabetes mellitus), 2) managing diabetes and risk factors to prevent diabetes-related complications (comprehensive lifestyle interventions; diabetes self-management education [DSME]; self-monitoring of blood glucose [SMBG]; intensive glycemic, blood pressure, and lipid control; integrated and coordinated care; smoking cessation), 3) screening for and early treatment of diabetes complications (cardiovascular disease [CVD], eye complications, foot ulcers, end-stage renal disease [ESRD]), or 4) treating diabetes-related complications and comorbidities (CVD, eye complications, foot ulcers, and comorbidities such as obesity, mental health, and sleep apnea). Within each of the four broader groups, we further classified studies into specific categories corresponding to each ADA-recommended intervention.
To facilitate comparisons across studies, we converted all costs and ICERs to 2017 U.S. dollars using the Consumer Price Index (11). If costs were reported in other currencies, we used the annual exchange rate from the Federal Reserve Bank (12) to convert them into U.S. dollars. If a study did not mention the year used in cost calculations, we assumed cost was for the one year prior to publication. ICERs were expressed as dollars per QALY or dollars per LYG and were rounded to the nearest hundred dollars. We calculated a range and median ICER for each intervention category.
Classifying the Interventions
We classified interventions based on their median levels of CE (five tiers) and strength of evidence (three levels). As there are no universally accepted thresholds to judge whether an intervention is cost-effective, we grouped the CE tiers of the intervention based on conventional norms according to their estimated ICERs: 1) cost-saving when the intervention generates better health outcomes and costs less than the comparison intervention or is cost neutral (ICER = 0), 2) very cost-effective if the ICER is greater than zero but less than or equal to $25,000 per QALY or LYG, 3) cost-effective if the ICER is greater than $25,000 but less than or equal to $50,000 per QALY or LYG, 4) marginally cost-effective if the ICER is greater than $50,000 but less than or equal to $100,000 per QALY or LYG, and 5) not cost-effective if the ICER is greater than $100,000 per QALY or LYG. Since there are no conventional norms for thresholds of cost per LYG, we used the same threshold for cost per QALY, which was the same approach used by Li et al. (4).
We classified the evidence level of the CE findings as strong, supportive, or uncertain. Strong evidence included findings from one of the following two categories:
Category 1:
The CE of the intervention was evaluated by two or more studies, AND
these studies were rated as at least “good” in quality, AND
the effectiveness of the interventions was based on either well-conducted, generalizable randomized clinical trials with adequate power or well-conducted meta-analyses or a diabetes disease simulation model that was validated, AND
the effectiveness of the intervention was also rated as level A or level B evidence by the ADA’s SOC 2019 (13), AND
the ICERs of the intervention from different studies consistently fell into the same CE tier.
Category 2:
CE assessment meets items 3 and 4 from category 1, but only one study evaluated the intervention, AND
the study was rated as “excellent” in quality.
We considered evidence on the CE of an intervention to be supportive if:
the CE of the intervention was evaluated by only one study AND this study was rated lower than “excellent” quality, OR
the effectiveness of the intervention was supported by level C evidence according to the ADA’s SOC 2019 (13) or by expert consensus only, OR
the CE was based on a simulation by a diabetes disease model that was not validated (a model was considered to be validated if it was explicitly stated in the text of the article or if the model was known to be validated based on previous literature).
For each ADA-recommended intervention within each evidence-CE level category, we described its comparison intervention, the study population in which the intervention was implemented, and the ADA’s level of evidence rating. We also reported the number of studies that evaluated the CE of this intervention (based on the current evidence and on the previous review), and the median and range of the ICERs across the studies.
Results
Our initial search yielded 18,195 articles over the period of June 2008 to July 2017. After removal of duplicates and screening of all abstracts, 1,445 articles remained, of which we reviewed full texts to identify 73 CE studies that met our inclusion criteria. We added the 49 relevant studies on diabetes management from the 2010 review by Li et al. (encompassing 1985–2008), bringing the total number of studies to 122 over the period 1985–2017 (Fig. 1). Table 1 describes the CE studies included in our final review by intervention category. Studies that evaluated multiple interventions or a single intervention in diverse subgroups were assigned to more than one intervention or population category, respectively.
Description of the CE studies for diabetes interventions
Source (author, year/country) . | Study population . | Intervention . | Comparison . | Study method . | Time horizon; discount rate . | ICER (in 2017 US$) . | |
---|---|---|---|---|---|---|---|
Screening for undiagnosed T2D | |||||||
Centers for Disease Control and Prevention, 1998/U.S. (21) | U.S. population aged 25 years and older | Opportunistic screening for undiagnosed T2D starting at age 25 years, then treatment (universal screening) | No screening and treatment until clinical diagnosis of T2D | Lifetime; 3% | $114,046/QALY | ||
Hoerger et al., 2004/U.S. (22) | Individuals with hypertension | Targeted screening for undiagnosed diabetes among persons with hypertension | No screening or treatment until clinical diagnosis of T2D | Lifetime; 3% | $59,436–$165,735/QALY, decreasing with age $89,535/QALY for age 45 years | ||
U.S. population | One-time opportunistic screening, then treatment (universal screening) | No screening or treatment until clinical diagnosis of T2D | $91,694–$240,157/QALY, decreasing with age $233,045/QALY for age 45 years | ||||
U.S. population | One-time opportunistic screening, then treatment (universal screening) | Targeted screening, then treatment | $273,812–$888,746/QALY increasing with age | ||||
Gillett et al., 2015/U.K. (23) | Adults aged 40–74 years with preDM and undiagnosed T2D | Prescreening with a risk score, then screening with A1C test (cutoff of 6%) | No screening | Computer simulation (Sheffield T2D Model) of LEADER study cohort | Lifetime; 5% | $2,088/QALY | |
Adults aged 40–74 years with preDM and undiagnosed T2D | Prescreening with a risk score, then screening with FPG test (cutoff of 5.5 mmol) | No screening | Computer simulation (Sheffield T2D Model) of LEADER study cohort | Lifetime; 1.5% | $4,301/QALY | ||
Kahn et al., 2010/U.S. (24) | U.S. population without DM, mean age 30 years | T2D screening every 3 years starting at age 30 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $14,807/QALY | |
T2D screening every 1 year starting at age 45 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $21,846/QALY | |||
T2D screening every 3 years starting at age 45 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $13,707/QALY | |||
T2D screening every 5 years starting at age 45 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $13,784/QALY | |||
T2D screening every 3 years starting at age 60 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $36,253/QALY | |||
T2D screening every 1 year following hypertension diagnosis | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $8,855/QALY | |||
T2D screening every 5 years following hypertension diagnosis | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $9,141/QALY | |||
T2D screening every 6 months starting at age 30 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $57,438/QALY | |||
Screening for and treating GDM | |||||||
Nicholson et al., 2005/U.S. (25) | 30-year-old pregnant women 24–28 weeks’ gestation | Sequential method (50-g GCT + 100-g GTT) | No screening or 75-g GTT | RCT | <1 year; 0% | Cost-saving | |
100-g GTT | No screening or 75-g GTT | Cost-saving | |||||
100-g GTT | Sequential method | $44,704/QALY for maternal outcomes, $11,430/QALY for neonatal outcomes | |||||
Werner et al., 2012/U.S. (26) | Simulated cohort of 100,000 pregnant women | Sequential method (50-g GCT at 24–28 weeks + 100-g GTT) [current practice] | No screening | Computer simulation (decision tree) | Lifetime; 3% | $19,746/QALY | |
Simulated cohort of 100,000 pregnant women | FPG at 1st prenatal visit + 75-g GTT at 24–28 weeks [practice proposed by IADPSG] | Sequential method (50-g GCT at 24–28 weeks + 100-g GTT) [current practice] | Computer simulation (decision tree) | Lifetime; 3% | $24,060/QALY | ||
Chen et al., 2016/Singapore (27) | Pregnant women at risk for GDM | Universal GDM screening (75-g OGTT) among all pregnant women | Targeted GDM screening among high risk women | Computer simulation (decision tree) | <1 year; 3% | $11,841/QALY | |
Targeted GDM screening | No screening | Computer simulation (decision tree) | <1 year; 3% | $10,047/QALY | |||
Danyliv et al., 2016/Ireland (28) | Pregnant women at risk for GDM | 75-g OGTT method in primary care setting, then treatment | No screening | Computer simulation (decision tree) | Lifetime; 5% | Cost-saving | |
75-g OGTT method in hospital setting, then treatment | No screening | Computer simulation (decision tree) | Lifetime; 5% | Cost-saving | |||
75-g OGTT method in hospital setting, then treatment | 75-g OGTT method in primary care setting, then treatment | Computer simulation (decision tree) | Lifetime; 5% | Cost-saving | |||
Intensive glycemic control | |||||||
DCCT Research Group, 1996/U.S. (29) | T1D | Intensive glycemic control through insulin management, self-monitoring, and outpatient visits. The goal was to achieve A1C level as normal as possible (6%) | Conventional therapy (less intensive) | DCCT multicenter RCT (n = 1,441) | Lifetime; 3% | $64,516/QALY | |
Eastman et al., 1997/U.S. (30) | Newly diagnosed T2D | Intensive treatment targeting maintenance of A1C level at 7.2% | Standard antidiabetic treatment targeting A1C level at 10% | DCCT (n = 1,441) | Lifetime; 3% | $22,098/QALY | |
Gray et al., 2000/U.K. (31) | T2D | Intensive insulin therapy through multiple insulin injections A1C <7% | Conventional management (mainly through diet) aiming at FPG<15 mmol/L | UKPDS multicenter RCT (n = 5,120) | 10 years; 6% | Cost-saving | |
Palmer et al., 2000/Switzerland (32) | T1D | Intensive insulin therapy | Conventional insulin therapy | Literature review | Lifetime; 3% | $59,182/LYG | |
Wake et al., 2000/Japan (33) | T2D | Intensive insulin therapy through multiple insulin injections A1C <7% | Conventional insulin therapy | Kumamoto study RCT (n = 110) | 10 years; 3% | Cost-saving | |
Clarke et al., 2001/U.K. (34) | Newly diagnosed T2D + overweight | Intensive blood glucose control with metformin aiming at FPG <6 mmol/L | Conventional treatment primarily with diet | UKPDS (n = 5,120) | Median 10.7 years; 6% | Cost-saving | |
Centers for Disease Control and Prevention, 2002/U.S. (35) | Newly diagnosed T2D | Intensive glycemic control with insulin or sulfonylurea aiming at FPG of 6 mmol/L | Conventional glucose control (mainly diet) | UKPDS (n = 5,120) | Lifetime; 3% | $78,740/QALY; increasing rapidly with age at diagnosis: $18,288/QALY for age 25–34 years; >$127,000–$3.9 million for age 55–94 years. Cost-saving under UKPDS cost scenario (no case management cost, much less self-testing, slightly fewer physician visits) | |
Scuffham and Carr, 2003/U.K. (36) | T1D | Continuous subcutaneous insulin intervention for persons using insulin pump | Multiple daily insulin injections | 1 systematic review, 1 meta-analysis | 8 years; 6% | $12,954/QALY | |
Roze et al., 2005/U.K. (37) | T1D | Continuous subcutaneous insulin infusion | Multiple daily insulin injections | DCCT (n = 1,441) mainly meta-analysis | 60 years; 3% | $23,495/QALY | |
Clarke et al., 2005/U.K. (38) | Newly diagnosed T2D requiring insulin | Intensive glycemic control with insulin or sulfonylurea at FPG <6 mmol/L | Conventional glucose control therapy (mainly diet) | UKPDS (n = 5,120) | Lifetime; 3.5% | $4,318/QALY | |
Newly diagnosed T2D + overweight | Intensive glycemic control with metformin | Conventional glucose control therapy (mainly diet) | Cost-saving | ||||
Eddy et al., 2005/U.S. (39) | Newly diagnosed T2D | Intensive DPP lifestyle with FPG >125 mmol/L; target: A1C level of 7% | Dietary advice | DPP (n = 3,234) | 30 years; 3% | $42,037/QALY | |
Cameron and Bennett, 2009/Canada (40) | T1D and T2D | Insulin aspart | Regular human insulin | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: Cost-saving | |
Among T2D: $28,261/QALY | |||||||
Insulin lispro | Regular human insulin | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: $36,440/QALY Among T2D: $164,460/QALY | |||
Insulin glargine | Insulin neutral protamine hagedorn | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: $110,506/QALY Among T2D: $808,061/QALY | |||
Insulin detemir | Insulin neutral protamine hagedorn | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: $487,266/QALY Among T2D: dominated (intervention was more costly, less effective) | |||
Howard et al., 2010/Australia (41) | T2D aged ≥25 years | Intensive glycemic control | Usual care | Markov computer simulation (AusDiab Study) | Lifetime; 5% | Cost-saving | |
Adults aged 50–69 years | Screening for T2D + intensive glycemic control | Usual care | Markov computer simulation (AusDiab Study) | Lifetime; 5% | $15,398/QALY | ||
Klarenbach et al., 2011/Canada (42) | T2D inadequately controlled by metformin | Metformin + sulfonylureas | Metformin alone | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | $14,094/QALY | |
Metformin + meglitinide | Metformin + sulfonylurea | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Metformin + α-glucosidase inhibitor | Metformin + sulfonylurea | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | $1,037,902/QALY | |||
Metformin + TZD | Metformin + α-glucosidase inhibitor | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | $5,106,028/QALY | |||
Metformin + DPP-4 inhibitor | Metformin + TZD | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Metformin + basal insulin | Metformin + TZD | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Metformin + biphasic insulin | Metformin + TZD | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Gordon et al., 2017/U.K. (43) | Adults with T2D (mean age 73 years) | Metformin + sulfonylurea | Metformin + DPP-4 inhibitor | Program evaluation/computer simulation (CORE Diabetes Model) | Lifetime; 3.5% | $30,264/QALY | |
Metformin + TZD | Metformin + DPP-4 inhibitor | Program evaluation/computer simulation (CORE Diabetes Model) | Lifetime; 3.5% | $24,857/QALY | |||
SMBG | |||||||
Simon et al., 2008/U.K. (44) | T2D, non–insulin treated | SMBG (less intensive) for 1 year | Standard care | Trial | 1 year; no discounting | Intervention associated with higher cost, worse outcome | |
SMBG (more intensive) for 1 year | Standard care | Trial | 1 year; no discounting | Intervention associated with higher cost, worse outcome | |||
Tunis and Minshall, 2008/U.S. (45) | T2D treated with oral agents in a large HMO | SBMG 1×/day | No SMBG | Kaiser Permanente longitudinal study of cohort of “new antidiabetic drug users” | 40 years; 3% | $10,414/QALY | |
SMBG 3×/day | No SMBG | 40 years; 3% | $8,763/QALY | ||||
SBMG 1×/day | No SMBG | 5 years | $30,734/QALY | ||||
10 years | $12,319/QALY | ||||||
SMBG 3×/day | No SMBG | 5 years | $38,481/QALY | ||||
10 years | $686/QALY | ||||||
Cameron et al., 2010/Canada (46) | T1D | SMBG | Standard care | Simulation (UKPDS Outcomes Model) | Lifetime; 5% | $138,669/QALY | |
Pollock et al., 2010/Switzerland (47) | T2D adults (mean age 63 years) on oral antidiabetics | SMBG 1×/day | Usual care | Computer simulation (CORE Diabetes Model) | 30 years; 3% | $10,341/QALY | |
SMBG 2×/day | Usual care | Computer simulation (CORE Diabetes Model) | 30 years; 3% | $14,568/QALY | |||
SMBG 3×/day | Usual care | Computer simulation (CORE Diabetes Model) | 30 years; 3% | $19,542/QALY | |||
Tunis and Minshall, 2010/U.S. (48) | T2D adults (mean age 60 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $36,916/QALY | |
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $26,160/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $35,828/QALY | |||
Tunis et al., 2010/France, Germany, Italy, Spain (49) | T2D adults in France (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $22,405/QALY | |
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $11,619/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $14,719/QALY | |||
T2D adults in Germany (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $3,020/QALY | ||
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $3,651/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $9,331/QALY | |||
T2D adults in Italy (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $23,478/QALY | ||
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $22,072/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $28,424/QALY | |||
T2D adults in Spain (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 6% | $6,771/QALY | ||
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 6% | $5,736/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 6% | $10,637/QALY | |||
Tunis, 2011/Canada (50) | T2D adults (mean age 60 years) not on insulin | Canadian Optimal Prescribing and Utilization Service (1.29 strips per day of self-monitored blood glucose) | No intervention | Computer simulation model | 40 years; 5% | $77,684/QALY | |
McQueen et al., 2015/Canada (51) | T1D adults (mean age 50 years) with baseline A1C 7.6% | Provision of SMBG device with strip price Can$0.73 and a 10% error (exceeding accuracy requirements by ISO) | Provision of SMBG device with strip price Can$0.60 and 15% error (accuracy meeting ISO) | Markov computer simulation model | Lifetime and 3 years; 5% | Lifetime: $130,820/QALY 3-year: cost-saving | |
Intensive hypertension control | |||||||
UKPDS Group, 1998/U.K. (52) | T2D + hypertension | Tight control of hypertension, BP <150/80 mmHg, ACEI, β-blocker, and other agents | Less tight control of BP (mmHg), initially <200/105, later 180/105 | UKPDS (n = 5,120) | Lifetime; 6% | Cost-saving | |
Elliot et al., 2000/U.S. (53) | T2D, hypertension, initially free of CVD or ESRD | Reduction of BP to 130/85 mmHg | Reduction of BP to 140/90 mmHg | Meta-analysis of data from epidemiological studies and clinical trials | Lifetime; 3% | ||
Treatment started at age 50 years | $1,524/LYG | ||||||
Treatment started at age 60 years | Cost-saving | ||||||
Treatment started at age 70 years | Cost-saving | ||||||
Centers for Disease Control and Prevention, 2002/U.S. (35) | T2D + hypertension | Intensified hypertension control (ACEI, β-blocker), average BP 144/82 mmHg | Moderate hypertension control, average BP 154/86 mmHg | UKPDS (n = 5,120) | Lifetime; 3% | Cost-saving | |
Clarke et al., 2005/U.K. (38) | T2D + hypertension | Tight BP control BP <150/85 mmHg, ACEI (captopril) or β-blocker (atenolol) | Less tight control of BP (mmHg), initial <200/105, later <180/105 | UKPDS (n = 5,120) | Lifetime; 3.5% | $254/QALY | |
Ly et al., 2009/U.S. (54) | Newly diagnosed T2D and existing hypertension | Hypertension management program for 1 year | Standard care | Markov computer simulation model | 1 year; costs discounted 3% | Cost-saving | |
Hypertension management program for 3 years | Standard care | Markov computer simulation model | 3 years; 3% | Cost-saving | |||
Hypertension management program for 5 years | Standard care | Markov computer simulation model | 5 years; 3% | Cost-saving | |||
Howard et al., 2010/Australia (41) | T2D (AusDiab) | ACEI treatment | Usual care | Markov computer simulation model | Lifetime; 5% | Cost-saving | |
Cholesterol control | |||||||
Herman et al., 1999/U.S. (55) | T2D + dyslipidemia + previous myocardial infarction or angina | Simvastatin | Placebo | RCT | 5 years; 3% for cost, 0% for benefit | Cost-saving | |
Jönsson et al., 1999/European countries (56) | T2D + dyslipidemia + previous myocardial infarction or angina | Simvastatin | Placebo | RCT | Lifetime; 3% | Cost-saving–$11,938/LYG in different countries. Median: $3,556/LYG | |
Grover et al., 2000/Canada (57) | T2D + dyslipidemia + CVD history, adults aged ≥60 years | Simvastatin | Placebo | RCT | Lifetime; 5% | $7,747–$15,621/LYG increasing with pretreatment of LDL cholesterol level. More cost-effective for men than women | |
Centers for Disease Control and Prevention, 2002/U.S. (35) | T2D + dyslipidemia, no CVD history | Pravastatin | Placebo | RCT | Lifetime; 3% | $98,806/QALY | |
Raikou et al., 2007/U.K. and Ireland (58) | T2D, no CVD history, no elevated LDL cholesterol, ≥1 CVD risk factor (retinopathy, microalbuminuria or macroalbuminuria, current smoking, or hypertension) | Atorvastatin | Placebo | RCT | Lifetime; 3% | $4,445/QALY | |
Sorensen et al., 2009/U.S. (59) | T2D adults (mean age 60 years) with T2D and mixed dyslipidemia | Maintaining lipid levels without particular targets, including through combination therapy as recommended by National Cholesterol Education Program Adult Treatment Panel III guidelines | Usual care | Computer simulation model | Lifetime; 3% | $67,873/QALY $70,291/CHD event avoided | |
de Vries et al., 2014/the Netherlands (60) | T2D (mean age 61.3 years) | Statin treatment started at time of T2D diagnosis | No lipid-regulating treatment | Markov computer simulation model | 10 years; costs discounted 4%, benefits discounted 1.5% | $3,294/QALY (<45 years: $84,012/QALY; ages 45–55: $12,174/QALY; 55–65 years: $3,640/QALY) | |
Smoking cessation | |||||||
Earnshaw et al., 2002/U.S. (61) | Newly diagnosed T2D + smoker, aged 25–84 years | Smoking cessation, standard antidiabetic care | Standard antidiabetic care | Lifetime; 3% | <$31,750/QALY | ||
Aged 85–94 years | $114,046/QALY | ||||||
Educational/DSME program | |||||||
Gozzoli et al., 2001/Switzerland (62) | T2D | Standard antidiabetic care plus educational program, self-monitoring, recommendations on diet and exercise, self-management of diabetes and complications, general health education | Standard antidiabetic care | Literature review | Lifetime; 3% | $5,080/LYG | |
Shearer et al., 2004/Germany (63) | T1D | Structured treatment and teaching program: educational course of training to self-manage diabetes and enjoy dietary freedom | Usual care (daily insulin injection) | RCT | Lifetime; 6% | Cost-saving | |
Brownson et al., 2009/U.S. (64) | Hispanic and African American adults with T2D; insured and uninsured | DM self-management program (DSME classes, walking clubs, group visits/classes, weekly phone follow-up, one-on-one self-management sessions, mental health services) provided by health care providers, community health educators, nurses in real-world setting for 3–4 years | No intervention (baseline treatment) | Simulation model using the CDC-RTI Diabetes Cost-Effectiveness Model | Lifetime (100 years max); 3% | $55,726/QALY | |
Gillett et al., 2010/U.K. (65) | Newly diagnosed T2D | DSME focused on lifestyle factors (diet + physical activity), facilitated by registered health care professionals trained as educators, 1 year | Standard care | Trial and computer simulation | Lifetime; 3.5% | Real-world cost data: $5,047/QALY Trial data: $12,994/QALY | |
Gillespie et al., 2014/Ireland (66) | T1D | Dose Adjustment for Normal Eating (DAFNE) program for 18 months; group-based structured education sessions on insulin dose adjustment, carbohydrate estimation, and hypoglycemia management | Usual care | Cluster randomized trial | 18 months; no discounting | Cost-saving | |
Kruger et al., 2013/U.K. (67) | Simulated cohort of patients with existing T1D (mean age 40 years) | Dose Adjustment for Normal Eating (DAFNE), a 5-day structured education program (flexible insulin therapy and insulin doses to match carbohydrate intake), delivered in groups of 6–8 | No intervention | Trial/Sheffield T1D Policy Simulation Model | Lifetime; 3.5% | $26,054/QALY | |
Gordon et al., 2014/Australia (68) | T2D adults | 24-week educational advice and feedback on DM self-management provided via weekly telephone calls + DM kit with handbook, glucose meter, test strips, cell phone | Standard care | Markov computer simulation model | 5 years; 5% | Cost-saving | |
Prezio et al., 2014/U.S. (69) | Uninsured Mexican American adults with T2D | One-to-one culturally tailored diabetes education and management program | Usual care | Computer simulation model | 20 years; 3% | 5-year duration: $114,354/QALY 10-year duration: $44,199/QALY 20-year duration: $405/QALY | |
Ryabov, 2014/U.S. (70) | Mexican American adults with T2D | Educational program following the National DPP and led by community health workers (monthly 40–60-min visits for 2 years) | Usual care | Computer simulation model | 5, 10, 20 years and lifetime; 3% | $17,964/QALY | |
Varney et al., 2016/Australia (71) | Adults (mean age 60 years) with poorly controlled T2D | Monthly tele-coaching by dietitian to address lifestyle modification, treatment adherence, goal setting, barriers to change | Usual care | Computer simulation model (UKPDS Outcomes Model) | 10 years; 5% | Cost-saving | |
Odnoletkova et al., 2016/Belgium (72) | T2D on glucose-lowering medication therapy | Telephone counseling intervention (SMBG, lifestyle, medications) delivered by diabetes nurse educators and consisting of five 30-min phone sessions over 6 months | Usual care | Markov computer simulation model | 40 years; costs discounted 3%, benefits discounted 1.5% | $8,238/QALY | |
Screening for and preventing diabetes complications | |||||||
Cardiovascular disease | |||||||
Li et al., 2010/U.S. (73) | T2D | Daily use of aspirin (80 mg) | No aspirin use | Computer simulation model | Lifetime; 3% | $7,646/LYG $2,395/QALY | |
van Giessen et al., 2016/the Netherlands (74) | T2D on oral drugs only and without previous diagnosis of heart failure | Screening for heart failure via EMR symptoms | No screening | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Men: $10,078/QALY Women: $10,413/QALY | |
Screening for heart failure via EMR symptoms and physical exam | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Intervention associated with higher cost and worse outcome | |||
Screening for heart failure via EMR symptoms and physical exam and natriuretic peptide | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Intervention associated with higher cost and worse outcome | |||
Screening for heart failure via EMR symptoms and physical exam and natriuretic peptide and ECG | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Intervention associated with higher cost and worse outcome | |||
Screening for heart failure via ECG | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Men: $47,963/QALY Women: $64,818/QALY | |||
Eye complications | |||||||
Javitt et al., 1994/U.S. (75) | Newly diagnosed T2D | Eight strategies for eye screening with dilation: screening every 1, 2, 3, or 4 years and more frequent follow-up screening for diabetes patients with background retinopathy | No screening | Cross-sectional/longitudinal studies | Lifetime; 5% | Cost-saving (all 8 strategies) | |
Javitt and Aiello, 1996/U.S. (76) | Newly diagnosed T1D and T2D | Annual eye screening with dilation for all patients with diabetes but no retinopathy | Eye screening in 60% of diabetes patients | Cross-sectional/longitudinal studies | Lifetime; 5% | $8,763/QALY | |
T1D | $5,461/QALY | ||||||
T2D | Examination every 6 months for those with retinopathy | $8,763/QALY | |||||
Palmer et al., 2000/Switzerland (32) | T1D | Annual eye screening and treatment, conventional insulin therapy | Conventional insulin therapy | Literature review | Lifetime; 3% | Cost-saving | |
Vijan et al., 2000/U.S. (77) | T2D | Eye screening for diabetes patients every 5 years; subsequent annual screening for those with background retinopathy | No screening | Epidemiological studies | Lifetime; 3% | $29,845/QALY | |
Eye screening for diabetes patients every 3 years; subsequent annual screening for those with background retinopathy | No screening | $34,290/QALY | |||||
Eye screening for diabetes patients every 2 years; subsequent annual screening for those with background retinopathy | No screening | $38,989/QALY | |||||
Eye screening for diabetes patients annually; subsequent annual screening for those with background retinopathy | No screening | $50,165/QALY | |||||
Eye screening for diabetes patients every 3 years | 5-year screening intervals | $41,656/QALY | |||||
Eye screening for diabetes patients every 2 years | 3-year screening intervals | $68,580/QALY | |||||
Annual eye screening for diabetes patients | 2-year screening intervals | $148,336/QALY | |||||
Maberley et al., 2003/Canada (78) | T1D and T2D | Screening using digital camera, with immediate assessment of quality or electronically transferred to a remote reading center | Retina specialists visit Moose Factory every 6 months to examine people with diabetes, and patients in outlying communities are flown to Moose Factory, Canada | 10 years; 5% | Cost-saving | ||
Kirkizlar et al., 2013/U.S. (79) | DM and diabetic retinopathy | Telemedicine for the screening of diabetic retinopathy | Usual care (diabetic retinopathy screening by an eye care professional) | Markov computer simulation model | Lifetime; 3% | Patient pool size: 3,000: $61,124/QALY 3,500: $53,013/QALY 4,000: $46,929/QALY 6,000: $32,735/QALY 9,000: $23,273/QALY Age (years): <30: −$16,313 (cost-saving) 30–39: −$12,599 (cost-saving) 40–49: −$7,320 (cost-saving) 50–59: $8,248 60–69: $16,800 70–79: $39,395 80–89: $87,975 90–99: $105,371 Race: Black or African American: $20,322 Native American: −$5,550 White: $24,779 Unanswered: $25,751 | |
Chan et al., 2015/Hong Kong (80) | Adults with DM | Annual screening and treatment for intermediate age-related macular degeneration | No screening | Markov computer simulation model | Lifetime (100 years max); 3% | $16,027/QALY | |
Kawasaki et al., 2015/Japan (81) | Adults with DM | Screening for diabetic retinopathy by ophthalmologists using dilated fundus examinations | No screening | Markov computer simulation model | 50 year; 3% | $13,533/QALY | |
Scanlon et al., 2015/U.K. (82) | DM | DR screening every 6 months | Annual DR screening | Decision-analytic model | Lifetime; 3.5% | $502,666/QALY | |
Annual DR screening | DR screening every 2 years | Decision-analytic model | Lifetime; 3.5% | $170,900/QALY | |||
DR screening every 2 years | DR screening every 3 years | Decision-analytic model | Lifetime; 3.5% | $79,599/QALY | |||
DR screening every 3 years | DR screening every 5 years | Decision-analytic model | Lifetime; 3.5% | $45,574/QALY | |||
Nguyen et al., 2017/Singapore (83) | T2D and diabetic retinopathy | Telemedicine-based DR screening program, with real-time assessment of DR photographs by a centralized team supported by tele-ophthalmology IT infrastructure | Usual care (family physician assessment of DR) | Computer simulation model (decision tree and Markov) | Lifetime; 3% | Cost-saving | |
Scotland et al., 2016/Scotland (84) | T1D and T2D | Annual DR screening for those with no or mild retinopathy and biannual screening for observable retinopathy/maculopathy | Screening at 2-year intervals for those with no DR at two consecutive screening episodes | Markov computer simulation model (continuous-time hidden) | 30 years; 3.5% | $404,733/QALY | |
Screening at 2-year interval for those observed with no DR at two consecutive screening episodes | Screening at 2-year interval for those with no DR and no DR previously recorded | Markov computer simulation model (continuous-time hidden) | 30 years; 3.5% | $836,344/QALY | |||
Screening at 2-year interval for those with no DR and no DR previously recorded | Screening at 2-year interval for those with no DR | Markov computer simulation model (continuous-time hidden) | 30 years; 3.5% | $128,865/QALY | |||
van Katwyk et al., 2017/Canada (85) | Existing DM | DR screening by optometrists are publicly insured | Usual care (DR screening by primary care physician or referral to ophthalmologists are publicly insured) | Computer simulation probabilistic decision-analytic model | 30 years; 5% | $1,399/QALY | |
Foot ulcers | |||||||
Ragnarson Tennvall and Apelqvist, 2001/Sweden (86) | T1D and T2D, moderate to high risk (previous foot ulcer/amputation, neuropathy) | Optimal prevention of foot ulcer including foot inspection, appropriate footwear, treatment, and education | Usual care | Clinical and epidemiological data | 5 years; 0% | Cost-saving | |
Low risk (no specific risk factor) | >$127,000/QALY | ||||||
Ortegon et al., 2004/the Netherlands (87) | Newly diagnosed T2D + foot ulcer | Intensive glycemic control + optimal foot care | Standard care | Trial | Lifetime; 3% | $57,023/QALY | |
End-stage renal disease | |||||||
Borch-Johnsen et al., 1993/Germany (88) | T1D | Annual screening for microalbuminuria at 5 years after diabetes onset + ACEI treatment | Treatment of macroalbuminuria | Cohort | 30 years; 6% | Cost-saving | |
Kiberd and Jindal, 1995/Canada (89) | T1D | Screening for microalbuminuria + ACEI treatment | Treatment of hypertension and/or macroproteinuria | Clinical trial | Lifetime; 5% | $74,168/QALY | |
Golan et al., 1999/U.S. (90) | Newly diagnosed T2D | Treat patients with new diagnosis with ACEI | Screening for macroalbuminuria and treatment with ACEI | RCT | Lifetime; 3% | Cost-saving | |
Screening for microalbuminuria and treatment with ACEI | Screening for macroalbuminuria and treatment with ACEI | Cost-saving | |||||
Treat patients with new diagnosis with ACEI | Screening for microalbuminuria and treatment with ACEI | $13,843/QALY | |||||
Clark et al., 2000/Canada (91) | T1D | Province or territory paying for ACEI | Pay from out of pocket | Collaborative observational study using admin data base | 21 years; 5% | Cost-saving | |
Palmer et al., 2000/Switzerland (32) | T1D, high cholesterol, high systolic BP | Microalbuminuria monitoring, ACE treatment, conventional insulin therapy | Conventional insulin therapy | Literature review | Lifetime; 3% | Cost-saving | |
Palmer et al., 2003/Belgium, France (92) | T2D + macroalbuminuria + hypertension | Irbesartan | Standard therapy for hypertension | RCT | Lifetime; 3% | Cost-saving | |
Souchet et al., 2003/France (93) | T2D + nephropathy | Losartan | Placebo | Trial | 4 years; costs discounted 8%, benefits not discounted | Cost-saving | |
Dong et al., 2004/U.S. (94) | T1D | ACEI treatment starting at 1 year after diagnosis | Annual screening for microalbuminuria ACE treatment | Trial | Lifetime; 3% | $48,260/QALY, increased with lowering A1C level, at A1C level 9%, <$31,750/QALY | |
Palmer et al., 2004/U.K. (95) | T2D + hypertension + nephropathy | Irbesartan | Standard therapy for hypertension | RCT | 10 years; 6% for costs, 1.5% for benefits | Cost-saving | |
Palmer et al., 2004/U.S. (96) | T2D + hypertension + microalbuminuria | Irbesartan | Standard therapy for hypertension | RCT | 25 years; 3% | Cost-saving | |
Szucs et al., 2004/Switzerland (97) | T2D + nephropathy | Losartan | Placebo | Trial | 3.5 years; 0% | Cost-saving | |
Palmer et al., 2005/Spain (98) | T2D + microalbuminuria + hypertension | Irbesartan | Standard therapy for hypertension, no ACEI, AIIRA, or β-blockers | RCT | 25 years; 3% | Cost-saving | |
Rosen et al., 2005/U.S. (99) | Medicare population (T1D and T2D) | Medicare full payment for ACEI (target: ACEI use increased by at least 7.2%) | Pay from out of pocket | RCT | Lifetime; 3% | Cost-saving | |
Coyle et al., 2007/Canada (100) | T2D + hypertension + macronephropathy + micronephropathy | Irbesartan added at stage of microalbuminuria | Conventional treatment for diabetes and hypertension, no ACEI or AIIRAs | RCT | Lifetime; 5% | Cost-saving | |
Palmer et al., 2007/Hungary (101) | T2D + microalbuminuria | Adding irbesartan | Placebo + standard therapy for hypertension | RCT | 25 years; 5% | Cost-saving | |
Palmer et al., 2007/U.K. (102) | T2D + hypertension + microalbuminuria | Irbesartan | Standard therapy for hypertension | RCT | 25 years; 3.5% | Cost-saving | |
Irbesartan added at stage of overt nephropathy | Conventional treatment for diabetes and hypertension | Cost-saving | |||||
Irbesartan added at stage of microalbuminuria | Irbesartan added at stage of overt nephropathy | Cost-saving | |||||
Howard et al., 2010/Australia (41) | Individuals aged 50–69 years with T2D from the AusDiab study | Screening for proteinuria + addition of an ACEI | Usual care | Markov computer simulation model | Lifetime; 5% | $5,310/QALY | |
Comprehensive interventions | |||||||
Palmer et al., 2000/Switzerland (32) | T1D | Conventional glycemic control + ACEI therapy + eye screening and treatment | Conventional glycemic control | Lifetime; 3% | Cost-saving | ||
Intensive insulin therapy + ACEI therapy | Intensive insulin therapy | $59,055/LYG | |||||
Intensive insulin therapy + eye screening | Intensive insulin therapy | $64,262/LYG | |||||
Intensive insulin therapy + ACEI therapy + eye screening | Intensive insulin therapy | $63,246/LYG | |||||
Gozzoli et al., 2001/Switzerland (62) | T2D | Added education program, nephropathy screening, and ACEI therapy to standard antidiabetic care | Standard antidiabetic care | Lifetime; 0%, 3% | Cost-saving | ||
Added education program, nephropathy screening, ACEI therapy, and retinopathy screening and laser therapy to standard antidiabetic care | Standard antidiabetic care | Cost-saving | |||||
Multifactorial intervention included educational program, screening for nephropathy and retinopathy, control of CVD risk factors, early diagnosis and treatment of complications, and health education | Standard antidiabetic care | Cost-saving | |||||
Gaede et al., 2008/Denmark (103) | T2D and microalbuminuria (mean age 55 years) | Intensive treatment for 7.8 years (stepwise implementation of behavior modification and pharmacologic therapy targeting hyperglycemia, hypertension, dyslipidemia, and microalbuminuria, and 2° prevention of CVD with aspirin) | Standard care | Markov computer simulation model | Lifetime; 3% | $4,629/QALY $7,162/LYG | |
Tasosa et al., 2010/U.S. (104) | Newly diagnosed T2D, African American adults | Aggressive hypertension control with ACEI or β-blocker, glycemic control with insulin or sulfonylurea, hyperlipidemia treatment based on pravastatin and four physician visits with blood/lipid/biochemical profiles | Usual care | Markov computer simulation model | Lifetime; 3% | $33,912/QALY | |
Newly diagnosed T2D | Aggressive hypertension control with ACEI or β-blocker, glycemic control with insulin or sulfonylurea, hyperlipidemia treatment based on pravastatin and four physician visits with blood/lipid/biochemical profiles | Usual care | Markov computer simulation model | Lifetime; 3% | $51,587/QALY | ||
Giorda et al., 2014/Italy (105) | T2D | Physician-led 5-year quality-of-care scheme to improve A1C, BP, lipids, and BMI | Standard care | Computer simulation model | 50 years; 3% | Cost-saving | |
Laxy et al., 2017/U.K. (106) | Newly diagnosed T2D (mean age 61.5 years) from ADDITION-UK | Intensive lifestyle changes and medication adherence, delivered by a specialist team of doctors, nurses, dietitians (2 years) | Usual care | Trial/UKPDS Outcomes model | 10, 20, and 30 years; 3.5% | 10-year: $98,613/QALY 20-year: $39,378/QALY 30-year: $38,139/QALY | |
Integrated and coordinated care | |||||||
Coordinated care | |||||||
Mason et al., 2005/England (107) | T2D + hypertension | Policy to implement clinics led by specialist nurses to treat and control hypertension through consultation, medication review, condition assessment, and lifestyle advice | Usual care | RCT | Lifetime; 5% | $6,096/QALY | |
Diagnosed diabetes + dyslipidemia | Policy to implement clinics led by specialist nurses to treat and control hyperlipidemia by usual care | Usual care | $29,972/QALY | ||||
Gilmer et al., 2007/U.S. (108) | Diabetes, 48% Latinos, uninsured population | Culturally sensitive case management and self-management training program led by bilingual/bicultural medical assistant and registered dietitian stepped-care pharmacologic management of glucose and lipid levels and hypertension | Standard care | Cohort study | 40 years; 3% | $15,240/QALY | |
McRae et al., 2008/Australia (109) | T2D | Integrated care program whereby GPs serve as case manager and program facilitates case management via provision of info and education to GPs (5 years) | Usual care | Computer simulation model | 40 years; 5% | $9,058/LYG $10,871/QALE | |
Schouten et al., 2010/the Netherlands (110) | Existing T2D | Integrated diabetes care with teams of 5–6 providers that attended learning sessions in quality-improvement techniques and diabetes care, and access to endocrinologists and diabetes educators for patients unresponsive to treatment or with difficult-to-manage diabetes. | Usual care | Computer simulation model (Dutch diabetes model) | Lifetime; costs discounted at 4.5%, benefits discounted at 1.5% | Men: $11,806/QALY Women: $13,474/QALY | |
Kuo et al., 2011/U.S. (111) | T2D patients at U.S. Air Force base | Diabetes management using the Chronic Care Model for 3 years | Usual care | Markov computer simulation model | 20 years; 3% | $55,465/QALY | |
Haji et al., 2013/Australia (112) | T2D | High level of practice nurse involvement in T2D management in primary care setting | Low level of practice nurse involvement in T2D management in primary care setting | Computer simulation model (UKPDS Outcomes Model) | 40 years; 5% | Cost-saving | |
Slingerland et al., 2013/the Netherlands (113) | T2D + A1C <7% | Patient-centered medical care in which patients receive detailed “diabetes passports” based on national guidelines for 1 year | Usual care | Trial | Lifetime; costs discounted 3% | Intervention was associated with higher costs and fewer QALYs | |
T2D + A1C 7–8.5% | Patient-centered medical care in which patients receive detailed “diabetes passports” based on national guidelines for 1 year | Usual care | Trial | Lifetime | $23,764/QALY | ||
T2D + A1C >8.5% | Patient-centered medical care in which patients receive detailed “diabetes passports” based on national guidelines for 1 year | Usual care | Trial | Lifetime; costs discounted 3% | $7,622/QALY | ||
Yu et al., 2013/U.S. (114) | Existing T2D + A1C >7% | Addition of a pharmacist to patient's care (prescribed/adjusted medications, ordered laboratory work, ordered/administered immunizations, provided DM self-management education, and worked to optimize overall glycemic and cardiovascular care of patients) | Usual care (primary care physician only) | Markov computer simulation model | 10 years; costs discounted 3%, benefits discounted 5% | Cost-saving | |
Tsiachristas et al., 2014/the Netherlands (115) | T2D and Charlson comorbidity index 2.22 | DM management program consisting of personal coaching and motivational interviewing | DM management program consisting of lifestyle interventions, periodic discussion sessions between providers and patients | Program evaluation | Not reported | Cost-saving | |
Wilson et al., 2014/U.K. (116) | T2D | Intermediate care clinics for diabetes, in which diabetes specialist nurses worked closely with hospital-based specialist teams and community services (podiatry and dietetic services) to manage patients until risk factor control was achieved (18 months max) | Usual care | Trial | 18 months; no discounting | $13,552/QALY | |
Tao et al., 2015/U.K. (117) | Adults with screen-detected T2D | Intensive DM care (more frequent provider contact, interactive audit and feedback sessions, theory-based education materials, dietitian referrals, group programs) | Usual care | Computer simulation model | 30 years; 3.5% | $70,649/QALY | |
Hirsch et al., 2017/U.S. (118) | T2D + complications (average of 8 comorbidities) | Obtaining care in an endocrinologist-pharmacist collaborative practice (3 personalized 60-min visits over 6 months) | Usual PCP visits | Program evaluation; Archimedes computer simulation model (VA Health System) | 2, 5, and 10 years; 3% | Cost-saving | |
Cobden et al., 2010/U.S. (119) | Medicare adults with T2D and preexisting complications | Injectable insulin (human or analog), without adherence | Oral medications (metformin +/− sulfonylurea or TZD) without adherence | Markov computer simulation model | Lifetime (35 years max); 3% | $15,251/QALY | |
Injectable insulin (human or analog insulin), with adjustments for adherence | Oral medications (metformin +/− sulfonylurea or TZD), with adjustments for adherence | Markov computer simulation model | Lifetime (35 years max); 3% | $20,476/QALY | |||
Decision support | |||||||
Cleveringa et al., 2010/the Netherlands (120) | T2D | Diabetes care protocol, consisting of a diabetes consultation hour run by a practice nurse, a CDSS diagnostic and treatment algorithm based on Dutch T2D guidelines, a recall system, and a feedback at both practice and patient level every 3 months | Usual care | Computer microsimulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | $73,253/QALY $19,360/LYG | |
O'Reilly et al., 2012/Canada (121) | T2D | Computerized decision support system linked to EMR, shared between patients and physicians | Usual care | Computer simulation model (Ontario Diabetes Economic Model) | 40 years; 5% | $190,417/QALY $185,831/LYG | |
Olvey, 2014/U.S. (122) | DM and hypertension or high cholesterol | Patients spoke by phone to a Medication Management Center pharmacist who discussed ACEI/ARB and statin guidelines, and potential addition of those treatments based on final recommendation by the patient's physician | Patients received a letter listing current prescription info and advising to discuss treatments with their physician | Computer simulation model (decision tree and Monte Carlo) | 5 years; costs discounted 5%, benefits discounted 2.5% | $5,710/5-year treatment success | |
Peer support | |||||||
Gillespie et al., 2012/Ireland (123) | T2D | Group-based peer support in addition to standardized diabetes care for 2 years | Standard care | Computer simulation model | Lifetime; 3.5% | Cost-saving | |
Treatment of diabetes-related complications | |||||||
Cardiovascular disease | |||||||
Hlatky et al., 2009/U.S., Canada, Brazil, Mexico, Czech Republic, Austria (124) | T2D and CHD | Prompt coronary revascularization combined with intensive medical management for 4 years | Intensive medical management, with coronary revascularization at a later date if clinically indicated | Trial | Lifetime; costs discounted 3% | Within trial: control dominant (Lifetime: $810/LYG) | |
CABG with intensive medical management | Intensive medical management, with coronary revascularization at a later date if clinically indicated | Trial | Lifetime; costs discounted 3% | Within trial: control dominant Lifetime: $63,401/LYG | |||
Patients taking metformin or rosiglitazone or both for 4 years | Patients on insulin or sulfonylurea or both | Trial | Lifetime; costs discounted 3% | Within trial: $395,245/QALY Lifetime: $70,146/LYG | |||
Eye complications | |||||||
Sharma et al., 2001/U.S. (125) | Diabetic retinopathy (HMO) | Immediate vitrectomy for management of vitreous hemorrhage secondary to diabetic retinopathy | Deferral of vitrectomy | DRVS | Lifetime; 6% | $3,683/QALY | |
Mitchell et al., 2012/U.K. (126) | Existing DM and DME | Ranibizumab monotherapy | Laser photocoagulation | Markov computer simulation model (RESTORE Study) | 15 years; 3.5% | $45,264/QALY | |
Ranibizumab combined with laser therapy | Laser photocoagulation | Markov computer simulation model (RESTORE Study) | 15 years; 3.5% | $68,017/QALY | |||
Hutton et al., 2017/U.S. (127) | DM and proliferative diabetic retinopathy, with and without DME | Ranibizumab (0.5 mg) | Laser photocoagulation | Trial | 2 years; no discounting | With DME: $56,752/QALY Without DME: $677,108/QALY | |
Foot ulcers | |||||||
Habacher et al., 2007/Austria (128) | Newly diagnosed diabetic food ulcer | Intensified treatment by international consensus on diabetic foot care | Standard treatment | Retrospective of patient records | 15 years; 0–8% | Cost-saving | |
O'Connor et al., 2008/U.S. (129) | DM and painful diabetic peripheral neuropathy | Duloxetine 60 mg 1×/day | Desipramine 100 mg 1×/day | Computer simulation model (decision tree) | 3 months; no discounting | $67,188/QALY | |
Pregabalin 100 mg 1×/day | Desipramine 100 mg 1×/day | Computer simulation model (decision tree) | 3 months; no discounting | Intervention associated with higher cost, worse outcome | |||
Gabapentin 800 mg 1×/day | Desipramine 100 mg 1×/day | Computer simulation model (decision tree) | 3 months; no discounting | Intervention associated with higher cost, worse outcome | |||
Cheng et al., 2017/Australia (130) | Simulated cohort of existing DM and at high risk of developing foot ulcers | Optimal care for foot ulcers and patient education | Usual care | Markov computer simulation model | 5 years; 5% | Cost-saving | |
Addressing diabetes comorbidities | |||||||
Obesity | |||||||
Anselmino et al., 2009/Austria, Italy, Spain (131) | T2D and BMI >35 kg/m2 | Gastric banding surgery | Usual care | Computer simulation model (deterministic linear algorithm) | 5 years; 3.5% | Austria: (−$5,027)/QALY, cost-saving Italy: (−$1,945)/QALY cost-saving Spain: $2,558/QALY | |
Gastric bypass surgery | Usual care | Computer simulation model (deterministic linear algorithm) | 5 years; 3.5% | Austria: (−$2,542)/QALY, cost-saving Italy: (−$2,189)/QALY, cost-saving Spain: $4,680/QALY | |||
Ikramuddin et al., 2009/U.S. (132) | T2D and obesity | Gastric bypass surgery | Standard medical management | Computer simulation model (CORE Diabetes Model) | 35 years; 3% | $29,641/QALY $40,032/LYG | |
Keating et al., 2009/Australia (133) | T2D and obesity (class I and II) | Gastric band surgery + conventional therapy for 2 years | Conventional therapy for 2 years | Computer simulation model | Lifetime; 3% | Cost-saving | |
Hoerger et al., 2010/U.S. (134) | Newly diagnosed or existing T2D and BMI ≥35 kg/m2 | Gastric bypass/gastric banding surgery | Standard care | Computer simulation model | Lifetime; 3% | For newly diagnosed DM: $10,254/QALY for gastric bypass $16,115/QALY for gastric banding For existing DM: $17,580/QALY for gastric bypass $19,045/QALY for gastric banding | |
Pollock et al., 2013/U.K. (135) | T2D and obesity | Gastric banding surgery | Standard care | Computer simulation model (CORE Diabetes Model) | 40 years; 3.5% | $6,785/QALY | |
Borisenko et al., 2015/Sweden (136) | T2D and obesity | Bariatric surgery | No surgery | Decision-analytic model using Markov processes | Lifetime; 3% | Bariatric surgery becomes cost-effective after 2 years ($39,604/QALY) and cost-saving after 17 years | |
James et al., 2017/Australia (137) | Simulated cohort of 30-year-old Australian females with T2D and obesity | Gastric banding surgery | Usual care (pharmacotherapy, diet, exercise management) | Markov computer simulation model | Lifetime; 5% | Cost-saving | |
Gastric bypass surgery | Usual care (pharmacotherapy, diet, exercise management) | Markov computer simulation model | Lifetime; 5% | Cost-saving | |||
Sleeve gastrectomy surgery | Usual care (pharmacotherapy, diet, exercise management) | Markov computer simulation model | Lifetime; 5% | Cost-saving | |||
Wentworth et al., 2017/U.S. (138) | T2D and overweight | Gastric banding surgery | Usual care | Computer simulation model (UKPDS Outcomes Model) | 2 and 10 years; 3% | Within 2-year trial: $100,050/QALY 5-year simulation: $55,120/QALY 10-year simulation: $30,747/QALY 15-year simulation: $23,320/QALY | |
Mental health | |||||||
Katon et al., 2006/U.S. (139) | Depression + poorly controlled DM or CHD | Multicondition collaborative treatment program led by a physician-supervised registered nurse and including patient education to promote self-care for 2 years (TEAMCare) | Usual care | Trial | NA | Cost-saving | |
Johnson et al., 2016/Canada (140) | T2D + depressive symptoms (PHQ ≥10) | Screening for depression + enhanced care (follow-up with family physician) | Usual care | Trial | 1 year; no discounting | $91,270/QALY | |
Screening for depression + coordinated, collaborative care led by a nurse care manager, in consultation with psychiatrists/endocrinologists (adapted TEAMCare) | Usual care | Trial | 1 year; no discounting | $29,160/QALY | |||
Screening for depression + coordinated, collaborative care led by a nurse care manager, in consultation with psychiatrists/endocrinologists (adapted TEAMCare) | Screening for depression + enhanced care (follow-up with family physician) | Trial | 1 year; no discounting | $18,980/QALY | |||
Kearns et al., 2017/U.K. (141) | Simulated cohort of existing T2D | Collaborative care | Usual care | Computer simulation (discrete event) | Lifetime; 3.5% | $18,814/QALY | |
Improved opportunistic screening for depression | Usual care | Computer simulation (discrete event) | Lifetime; 3.5% | $111,180/QALY | |||
Collaborative care + improved opportunistic screening for depression | Usual care | Computer simulation (discrete event) | Lifetime; 3.5% | $65,201/QALY | |||
Sleep apnea | |||||||
Guest et al., 2014/U.K. (142) | T2D with obstructive sleep apnea | Treatment with CPAP for 5 years | Standard care | Program evaluation/trial | 5 years; no discounting | $27,750/QALY |
Source (author, year/country) . | Study population . | Intervention . | Comparison . | Study method . | Time horizon; discount rate . | ICER (in 2017 US$) . | |
---|---|---|---|---|---|---|---|
Screening for undiagnosed T2D | |||||||
Centers for Disease Control and Prevention, 1998/U.S. (21) | U.S. population aged 25 years and older | Opportunistic screening for undiagnosed T2D starting at age 25 years, then treatment (universal screening) | No screening and treatment until clinical diagnosis of T2D | Lifetime; 3% | $114,046/QALY | ||
Hoerger et al., 2004/U.S. (22) | Individuals with hypertension | Targeted screening for undiagnosed diabetes among persons with hypertension | No screening or treatment until clinical diagnosis of T2D | Lifetime; 3% | $59,436–$165,735/QALY, decreasing with age $89,535/QALY for age 45 years | ||
U.S. population | One-time opportunistic screening, then treatment (universal screening) | No screening or treatment until clinical diagnosis of T2D | $91,694–$240,157/QALY, decreasing with age $233,045/QALY for age 45 years | ||||
U.S. population | One-time opportunistic screening, then treatment (universal screening) | Targeted screening, then treatment | $273,812–$888,746/QALY increasing with age | ||||
Gillett et al., 2015/U.K. (23) | Adults aged 40–74 years with preDM and undiagnosed T2D | Prescreening with a risk score, then screening with A1C test (cutoff of 6%) | No screening | Computer simulation (Sheffield T2D Model) of LEADER study cohort | Lifetime; 5% | $2,088/QALY | |
Adults aged 40–74 years with preDM and undiagnosed T2D | Prescreening with a risk score, then screening with FPG test (cutoff of 5.5 mmol) | No screening | Computer simulation (Sheffield T2D Model) of LEADER study cohort | Lifetime; 1.5% | $4,301/QALY | ||
Kahn et al., 2010/U.S. (24) | U.S. population without DM, mean age 30 years | T2D screening every 3 years starting at age 30 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $14,807/QALY | |
T2D screening every 1 year starting at age 45 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $21,846/QALY | |||
T2D screening every 3 years starting at age 45 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $13,707/QALY | |||
T2D screening every 5 years starting at age 45 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $13,784/QALY | |||
T2D screening every 3 years starting at age 60 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $36,253/QALY | |||
T2D screening every 1 year following hypertension diagnosis | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $8,855/QALY | |||
T2D screening every 5 years following hypertension diagnosis | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $9,141/QALY | |||
T2D screening every 6 months starting at age 30 years | No screening | Computer simulation (Archimedes) | Lifetime; 3% | $57,438/QALY | |||
Screening for and treating GDM | |||||||
Nicholson et al., 2005/U.S. (25) | 30-year-old pregnant women 24–28 weeks’ gestation | Sequential method (50-g GCT + 100-g GTT) | No screening or 75-g GTT | RCT | <1 year; 0% | Cost-saving | |
100-g GTT | No screening or 75-g GTT | Cost-saving | |||||
100-g GTT | Sequential method | $44,704/QALY for maternal outcomes, $11,430/QALY for neonatal outcomes | |||||
Werner et al., 2012/U.S. (26) | Simulated cohort of 100,000 pregnant women | Sequential method (50-g GCT at 24–28 weeks + 100-g GTT) [current practice] | No screening | Computer simulation (decision tree) | Lifetime; 3% | $19,746/QALY | |
Simulated cohort of 100,000 pregnant women | FPG at 1st prenatal visit + 75-g GTT at 24–28 weeks [practice proposed by IADPSG] | Sequential method (50-g GCT at 24–28 weeks + 100-g GTT) [current practice] | Computer simulation (decision tree) | Lifetime; 3% | $24,060/QALY | ||
Chen et al., 2016/Singapore (27) | Pregnant women at risk for GDM | Universal GDM screening (75-g OGTT) among all pregnant women | Targeted GDM screening among high risk women | Computer simulation (decision tree) | <1 year; 3% | $11,841/QALY | |
Targeted GDM screening | No screening | Computer simulation (decision tree) | <1 year; 3% | $10,047/QALY | |||
Danyliv et al., 2016/Ireland (28) | Pregnant women at risk for GDM | 75-g OGTT method in primary care setting, then treatment | No screening | Computer simulation (decision tree) | Lifetime; 5% | Cost-saving | |
75-g OGTT method in hospital setting, then treatment | No screening | Computer simulation (decision tree) | Lifetime; 5% | Cost-saving | |||
75-g OGTT method in hospital setting, then treatment | 75-g OGTT method in primary care setting, then treatment | Computer simulation (decision tree) | Lifetime; 5% | Cost-saving | |||
Intensive glycemic control | |||||||
DCCT Research Group, 1996/U.S. (29) | T1D | Intensive glycemic control through insulin management, self-monitoring, and outpatient visits. The goal was to achieve A1C level as normal as possible (6%) | Conventional therapy (less intensive) | DCCT multicenter RCT (n = 1,441) | Lifetime; 3% | $64,516/QALY | |
Eastman et al., 1997/U.S. (30) | Newly diagnosed T2D | Intensive treatment targeting maintenance of A1C level at 7.2% | Standard antidiabetic treatment targeting A1C level at 10% | DCCT (n = 1,441) | Lifetime; 3% | $22,098/QALY | |
Gray et al., 2000/U.K. (31) | T2D | Intensive insulin therapy through multiple insulin injections A1C <7% | Conventional management (mainly through diet) aiming at FPG<15 mmol/L | UKPDS multicenter RCT (n = 5,120) | 10 years; 6% | Cost-saving | |
Palmer et al., 2000/Switzerland (32) | T1D | Intensive insulin therapy | Conventional insulin therapy | Literature review | Lifetime; 3% | $59,182/LYG | |
Wake et al., 2000/Japan (33) | T2D | Intensive insulin therapy through multiple insulin injections A1C <7% | Conventional insulin therapy | Kumamoto study RCT (n = 110) | 10 years; 3% | Cost-saving | |
Clarke et al., 2001/U.K. (34) | Newly diagnosed T2D + overweight | Intensive blood glucose control with metformin aiming at FPG <6 mmol/L | Conventional treatment primarily with diet | UKPDS (n = 5,120) | Median 10.7 years; 6% | Cost-saving | |
Centers for Disease Control and Prevention, 2002/U.S. (35) | Newly diagnosed T2D | Intensive glycemic control with insulin or sulfonylurea aiming at FPG of 6 mmol/L | Conventional glucose control (mainly diet) | UKPDS (n = 5,120) | Lifetime; 3% | $78,740/QALY; increasing rapidly with age at diagnosis: $18,288/QALY for age 25–34 years; >$127,000–$3.9 million for age 55–94 years. Cost-saving under UKPDS cost scenario (no case management cost, much less self-testing, slightly fewer physician visits) | |
Scuffham and Carr, 2003/U.K. (36) | T1D | Continuous subcutaneous insulin intervention for persons using insulin pump | Multiple daily insulin injections | 1 systematic review, 1 meta-analysis | 8 years; 6% | $12,954/QALY | |
Roze et al., 2005/U.K. (37) | T1D | Continuous subcutaneous insulin infusion | Multiple daily insulin injections | DCCT (n = 1,441) mainly meta-analysis | 60 years; 3% | $23,495/QALY | |
Clarke et al., 2005/U.K. (38) | Newly diagnosed T2D requiring insulin | Intensive glycemic control with insulin or sulfonylurea at FPG <6 mmol/L | Conventional glucose control therapy (mainly diet) | UKPDS (n = 5,120) | Lifetime; 3.5% | $4,318/QALY | |
Newly diagnosed T2D + overweight | Intensive glycemic control with metformin | Conventional glucose control therapy (mainly diet) | Cost-saving | ||||
Eddy et al., 2005/U.S. (39) | Newly diagnosed T2D | Intensive DPP lifestyle with FPG >125 mmol/L; target: A1C level of 7% | Dietary advice | DPP (n = 3,234) | 30 years; 3% | $42,037/QALY | |
Cameron and Bennett, 2009/Canada (40) | T1D and T2D | Insulin aspart | Regular human insulin | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: Cost-saving | |
Among T2D: $28,261/QALY | |||||||
Insulin lispro | Regular human insulin | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: $36,440/QALY Among T2D: $164,460/QALY | |||
Insulin glargine | Insulin neutral protamine hagedorn | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: $110,506/QALY Among T2D: $808,061/QALY | |||
Insulin detemir | Insulin neutral protamine hagedorn | Computer simulation (Center for Outcomes Research Model) | 35 and 60 years; 5% | Among T1D: $487,266/QALY Among T2D: dominated (intervention was more costly, less effective) | |||
Howard et al., 2010/Australia (41) | T2D aged ≥25 years | Intensive glycemic control | Usual care | Markov computer simulation (AusDiab Study) | Lifetime; 5% | Cost-saving | |
Adults aged 50–69 years | Screening for T2D + intensive glycemic control | Usual care | Markov computer simulation (AusDiab Study) | Lifetime; 5% | $15,398/QALY | ||
Klarenbach et al., 2011/Canada (42) | T2D inadequately controlled by metformin | Metformin + sulfonylureas | Metformin alone | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | $14,094/QALY | |
Metformin + meglitinide | Metformin + sulfonylurea | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Metformin + α-glucosidase inhibitor | Metformin + sulfonylurea | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | $1,037,902/QALY | |||
Metformin + TZD | Metformin + α-glucosidase inhibitor | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | $5,106,028/QALY | |||
Metformin + DPP-4 inhibitor | Metformin + TZD | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Metformin + basal insulin | Metformin + TZD | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Metformin + biphasic insulin | Metformin + TZD | Computer simulation (UKPDS Outcomes Model) | Lifetime; 5% | Intervention associated with higher cost, worse outcome | |||
Gordon et al., 2017/U.K. (43) | Adults with T2D (mean age 73 years) | Metformin + sulfonylurea | Metformin + DPP-4 inhibitor | Program evaluation/computer simulation (CORE Diabetes Model) | Lifetime; 3.5% | $30,264/QALY | |
Metformin + TZD | Metformin + DPP-4 inhibitor | Program evaluation/computer simulation (CORE Diabetes Model) | Lifetime; 3.5% | $24,857/QALY | |||
SMBG | |||||||
Simon et al., 2008/U.K. (44) | T2D, non–insulin treated | SMBG (less intensive) for 1 year | Standard care | Trial | 1 year; no discounting | Intervention associated with higher cost, worse outcome | |
SMBG (more intensive) for 1 year | Standard care | Trial | 1 year; no discounting | Intervention associated with higher cost, worse outcome | |||
Tunis and Minshall, 2008/U.S. (45) | T2D treated with oral agents in a large HMO | SBMG 1×/day | No SMBG | Kaiser Permanente longitudinal study of cohort of “new antidiabetic drug users” | 40 years; 3% | $10,414/QALY | |
SMBG 3×/day | No SMBG | 40 years; 3% | $8,763/QALY | ||||
SBMG 1×/day | No SMBG | 5 years | $30,734/QALY | ||||
10 years | $12,319/QALY | ||||||
SMBG 3×/day | No SMBG | 5 years | $38,481/QALY | ||||
10 years | $686/QALY | ||||||
Cameron et al., 2010/Canada (46) | T1D | SMBG | Standard care | Simulation (UKPDS Outcomes Model) | Lifetime; 5% | $138,669/QALY | |
Pollock et al., 2010/Switzerland (47) | T2D adults (mean age 63 years) on oral antidiabetics | SMBG 1×/day | Usual care | Computer simulation (CORE Diabetes Model) | 30 years; 3% | $10,341/QALY | |
SMBG 2×/day | Usual care | Computer simulation (CORE Diabetes Model) | 30 years; 3% | $14,568/QALY | |||
SMBG 3×/day | Usual care | Computer simulation (CORE Diabetes Model) | 30 years; 3% | $19,542/QALY | |||
Tunis and Minshall, 2010/U.S. (48) | T2D adults (mean age 60 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $36,916/QALY | |
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $26,160/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $35,828/QALY | |||
Tunis et al., 2010/France, Germany, Italy, Spain (49) | T2D adults in France (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $22,405/QALY | |
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $11,619/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $14,719/QALY | |||
T2D adults in Germany (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $3,020/QALY | ||
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $3,651/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $9,331/QALY | |||
T2D adults in Italy (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 3% | $23,478/QALY | ||
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 3% | $22,072/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 3% | $28,424/QALY | |||
T2D adults in Spain (mean age 63 years) on oral treatment | SMBG 1×/day | No intervention | Computer simulation model | 40 years; 6% | $6,771/QALY | ||
SMBG 2×/day | No intervention | Computer simulation model | 40 years; 6% | $5,736/QALY | |||
SMBG 3×/day | No intervention | Computer simulation model | 40 years; 6% | $10,637/QALY | |||
Tunis, 2011/Canada (50) | T2D adults (mean age 60 years) not on insulin | Canadian Optimal Prescribing and Utilization Service (1.29 strips per day of self-monitored blood glucose) | No intervention | Computer simulation model | 40 years; 5% | $77,684/QALY | |
McQueen et al., 2015/Canada (51) | T1D adults (mean age 50 years) with baseline A1C 7.6% | Provision of SMBG device with strip price Can$0.73 and a 10% error (exceeding accuracy requirements by ISO) | Provision of SMBG device with strip price Can$0.60 and 15% error (accuracy meeting ISO) | Markov computer simulation model | Lifetime and 3 years; 5% | Lifetime: $130,820/QALY 3-year: cost-saving | |
Intensive hypertension control | |||||||
UKPDS Group, 1998/U.K. (52) | T2D + hypertension | Tight control of hypertension, BP <150/80 mmHg, ACEI, β-blocker, and other agents | Less tight control of BP (mmHg), initially <200/105, later 180/105 | UKPDS (n = 5,120) | Lifetime; 6% | Cost-saving | |
Elliot et al., 2000/U.S. (53) | T2D, hypertension, initially free of CVD or ESRD | Reduction of BP to 130/85 mmHg | Reduction of BP to 140/90 mmHg | Meta-analysis of data from epidemiological studies and clinical trials | Lifetime; 3% | ||
Treatment started at age 50 years | $1,524/LYG | ||||||
Treatment started at age 60 years | Cost-saving | ||||||
Treatment started at age 70 years | Cost-saving | ||||||
Centers for Disease Control and Prevention, 2002/U.S. (35) | T2D + hypertension | Intensified hypertension control (ACEI, β-blocker), average BP 144/82 mmHg | Moderate hypertension control, average BP 154/86 mmHg | UKPDS (n = 5,120) | Lifetime; 3% | Cost-saving | |
Clarke et al., 2005/U.K. (38) | T2D + hypertension | Tight BP control BP <150/85 mmHg, ACEI (captopril) or β-blocker (atenolol) | Less tight control of BP (mmHg), initial <200/105, later <180/105 | UKPDS (n = 5,120) | Lifetime; 3.5% | $254/QALY | |
Ly et al., 2009/U.S. (54) | Newly diagnosed T2D and existing hypertension | Hypertension management program for 1 year | Standard care | Markov computer simulation model | 1 year; costs discounted 3% | Cost-saving | |
Hypertension management program for 3 years | Standard care | Markov computer simulation model | 3 years; 3% | Cost-saving | |||
Hypertension management program for 5 years | Standard care | Markov computer simulation model | 5 years; 3% | Cost-saving | |||
Howard et al., 2010/Australia (41) | T2D (AusDiab) | ACEI treatment | Usual care | Markov computer simulation model | Lifetime; 5% | Cost-saving | |
Cholesterol control | |||||||
Herman et al., 1999/U.S. (55) | T2D + dyslipidemia + previous myocardial infarction or angina | Simvastatin | Placebo | RCT | 5 years; 3% for cost, 0% for benefit | Cost-saving | |
Jönsson et al., 1999/European countries (56) | T2D + dyslipidemia + previous myocardial infarction or angina | Simvastatin | Placebo | RCT | Lifetime; 3% | Cost-saving–$11,938/LYG in different countries. Median: $3,556/LYG | |
Grover et al., 2000/Canada (57) | T2D + dyslipidemia + CVD history, adults aged ≥60 years | Simvastatin | Placebo | RCT | Lifetime; 5% | $7,747–$15,621/LYG increasing with pretreatment of LDL cholesterol level. More cost-effective for men than women | |
Centers for Disease Control and Prevention, 2002/U.S. (35) | T2D + dyslipidemia, no CVD history | Pravastatin | Placebo | RCT | Lifetime; 3% | $98,806/QALY | |
Raikou et al., 2007/U.K. and Ireland (58) | T2D, no CVD history, no elevated LDL cholesterol, ≥1 CVD risk factor (retinopathy, microalbuminuria or macroalbuminuria, current smoking, or hypertension) | Atorvastatin | Placebo | RCT | Lifetime; 3% | $4,445/QALY | |
Sorensen et al., 2009/U.S. (59) | T2D adults (mean age 60 years) with T2D and mixed dyslipidemia | Maintaining lipid levels without particular targets, including through combination therapy as recommended by National Cholesterol Education Program Adult Treatment Panel III guidelines | Usual care | Computer simulation model | Lifetime; 3% | $67,873/QALY $70,291/CHD event avoided | |
de Vries et al., 2014/the Netherlands (60) | T2D (mean age 61.3 years) | Statin treatment started at time of T2D diagnosis | No lipid-regulating treatment | Markov computer simulation model | 10 years; costs discounted 4%, benefits discounted 1.5% | $3,294/QALY (<45 years: $84,012/QALY; ages 45–55: $12,174/QALY; 55–65 years: $3,640/QALY) | |
Smoking cessation | |||||||
Earnshaw et al., 2002/U.S. (61) | Newly diagnosed T2D + smoker, aged 25–84 years | Smoking cessation, standard antidiabetic care | Standard antidiabetic care | Lifetime; 3% | <$31,750/QALY | ||
Aged 85–94 years | $114,046/QALY | ||||||
Educational/DSME program | |||||||
Gozzoli et al., 2001/Switzerland (62) | T2D | Standard antidiabetic care plus educational program, self-monitoring, recommendations on diet and exercise, self-management of diabetes and complications, general health education | Standard antidiabetic care | Literature review | Lifetime; 3% | $5,080/LYG | |
Shearer et al., 2004/Germany (63) | T1D | Structured treatment and teaching program: educational course of training to self-manage diabetes and enjoy dietary freedom | Usual care (daily insulin injection) | RCT | Lifetime; 6% | Cost-saving | |
Brownson et al., 2009/U.S. (64) | Hispanic and African American adults with T2D; insured and uninsured | DM self-management program (DSME classes, walking clubs, group visits/classes, weekly phone follow-up, one-on-one self-management sessions, mental health services) provided by health care providers, community health educators, nurses in real-world setting for 3–4 years | No intervention (baseline treatment) | Simulation model using the CDC-RTI Diabetes Cost-Effectiveness Model | Lifetime (100 years max); 3% | $55,726/QALY | |
Gillett et al., 2010/U.K. (65) | Newly diagnosed T2D | DSME focused on lifestyle factors (diet + physical activity), facilitated by registered health care professionals trained as educators, 1 year | Standard care | Trial and computer simulation | Lifetime; 3.5% | Real-world cost data: $5,047/QALY Trial data: $12,994/QALY | |
Gillespie et al., 2014/Ireland (66) | T1D | Dose Adjustment for Normal Eating (DAFNE) program for 18 months; group-based structured education sessions on insulin dose adjustment, carbohydrate estimation, and hypoglycemia management | Usual care | Cluster randomized trial | 18 months; no discounting | Cost-saving | |
Kruger et al., 2013/U.K. (67) | Simulated cohort of patients with existing T1D (mean age 40 years) | Dose Adjustment for Normal Eating (DAFNE), a 5-day structured education program (flexible insulin therapy and insulin doses to match carbohydrate intake), delivered in groups of 6–8 | No intervention | Trial/Sheffield T1D Policy Simulation Model | Lifetime; 3.5% | $26,054/QALY | |
Gordon et al., 2014/Australia (68) | T2D adults | 24-week educational advice and feedback on DM self-management provided via weekly telephone calls + DM kit with handbook, glucose meter, test strips, cell phone | Standard care | Markov computer simulation model | 5 years; 5% | Cost-saving | |
Prezio et al., 2014/U.S. (69) | Uninsured Mexican American adults with T2D | One-to-one culturally tailored diabetes education and management program | Usual care | Computer simulation model | 20 years; 3% | 5-year duration: $114,354/QALY 10-year duration: $44,199/QALY 20-year duration: $405/QALY | |
Ryabov, 2014/U.S. (70) | Mexican American adults with T2D | Educational program following the National DPP and led by community health workers (monthly 40–60-min visits for 2 years) | Usual care | Computer simulation model | 5, 10, 20 years and lifetime; 3% | $17,964/QALY | |
Varney et al., 2016/Australia (71) | Adults (mean age 60 years) with poorly controlled T2D | Monthly tele-coaching by dietitian to address lifestyle modification, treatment adherence, goal setting, barriers to change | Usual care | Computer simulation model (UKPDS Outcomes Model) | 10 years; 5% | Cost-saving | |
Odnoletkova et al., 2016/Belgium (72) | T2D on glucose-lowering medication therapy | Telephone counseling intervention (SMBG, lifestyle, medications) delivered by diabetes nurse educators and consisting of five 30-min phone sessions over 6 months | Usual care | Markov computer simulation model | 40 years; costs discounted 3%, benefits discounted 1.5% | $8,238/QALY | |
Screening for and preventing diabetes complications | |||||||
Cardiovascular disease | |||||||
Li et al., 2010/U.S. (73) | T2D | Daily use of aspirin (80 mg) | No aspirin use | Computer simulation model | Lifetime; 3% | $7,646/LYG $2,395/QALY | |
van Giessen et al., 2016/the Netherlands (74) | T2D on oral drugs only and without previous diagnosis of heart failure | Screening for heart failure via EMR symptoms | No screening | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Men: $10,078/QALY Women: $10,413/QALY | |
Screening for heart failure via EMR symptoms and physical exam | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Intervention associated with higher cost and worse outcome | |||
Screening for heart failure via EMR symptoms and physical exam and natriuretic peptide | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Intervention associated with higher cost and worse outcome | |||
Screening for heart failure via EMR symptoms and physical exam and natriuretic peptide and ECG | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Intervention associated with higher cost and worse outcome | |||
Screening for heart failure via ECG | Screening for heart failure via EMR symptoms | Markov computer simulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | Men: $47,963/QALY Women: $64,818/QALY | |||
Eye complications | |||||||
Javitt et al., 1994/U.S. (75) | Newly diagnosed T2D | Eight strategies for eye screening with dilation: screening every 1, 2, 3, or 4 years and more frequent follow-up screening for diabetes patients with background retinopathy | No screening | Cross-sectional/longitudinal studies | Lifetime; 5% | Cost-saving (all 8 strategies) | |
Javitt and Aiello, 1996/U.S. (76) | Newly diagnosed T1D and T2D | Annual eye screening with dilation for all patients with diabetes but no retinopathy | Eye screening in 60% of diabetes patients | Cross-sectional/longitudinal studies | Lifetime; 5% | $8,763/QALY | |
T1D | $5,461/QALY | ||||||
T2D | Examination every 6 months for those with retinopathy | $8,763/QALY | |||||
Palmer et al., 2000/Switzerland (32) | T1D | Annual eye screening and treatment, conventional insulin therapy | Conventional insulin therapy | Literature review | Lifetime; 3% | Cost-saving | |
Vijan et al., 2000/U.S. (77) | T2D | Eye screening for diabetes patients every 5 years; subsequent annual screening for those with background retinopathy | No screening | Epidemiological studies | Lifetime; 3% | $29,845/QALY | |
Eye screening for diabetes patients every 3 years; subsequent annual screening for those with background retinopathy | No screening | $34,290/QALY | |||||
Eye screening for diabetes patients every 2 years; subsequent annual screening for those with background retinopathy | No screening | $38,989/QALY | |||||
Eye screening for diabetes patients annually; subsequent annual screening for those with background retinopathy | No screening | $50,165/QALY | |||||
Eye screening for diabetes patients every 3 years | 5-year screening intervals | $41,656/QALY | |||||
Eye screening for diabetes patients every 2 years | 3-year screening intervals | $68,580/QALY | |||||
Annual eye screening for diabetes patients | 2-year screening intervals | $148,336/QALY | |||||
Maberley et al., 2003/Canada (78) | T1D and T2D | Screening using digital camera, with immediate assessment of quality or electronically transferred to a remote reading center | Retina specialists visit Moose Factory every 6 months to examine people with diabetes, and patients in outlying communities are flown to Moose Factory, Canada | 10 years; 5% | Cost-saving | ||
Kirkizlar et al., 2013/U.S. (79) | DM and diabetic retinopathy | Telemedicine for the screening of diabetic retinopathy | Usual care (diabetic retinopathy screening by an eye care professional) | Markov computer simulation model | Lifetime; 3% | Patient pool size: 3,000: $61,124/QALY 3,500: $53,013/QALY 4,000: $46,929/QALY 6,000: $32,735/QALY 9,000: $23,273/QALY Age (years): <30: −$16,313 (cost-saving) 30–39: −$12,599 (cost-saving) 40–49: −$7,320 (cost-saving) 50–59: $8,248 60–69: $16,800 70–79: $39,395 80–89: $87,975 90–99: $105,371 Race: Black or African American: $20,322 Native American: −$5,550 White: $24,779 Unanswered: $25,751 | |
Chan et al., 2015/Hong Kong (80) | Adults with DM | Annual screening and treatment for intermediate age-related macular degeneration | No screening | Markov computer simulation model | Lifetime (100 years max); 3% | $16,027/QALY | |
Kawasaki et al., 2015/Japan (81) | Adults with DM | Screening for diabetic retinopathy by ophthalmologists using dilated fundus examinations | No screening | Markov computer simulation model | 50 year; 3% | $13,533/QALY | |
Scanlon et al., 2015/U.K. (82) | DM | DR screening every 6 months | Annual DR screening | Decision-analytic model | Lifetime; 3.5% | $502,666/QALY | |
Annual DR screening | DR screening every 2 years | Decision-analytic model | Lifetime; 3.5% | $170,900/QALY | |||
DR screening every 2 years | DR screening every 3 years | Decision-analytic model | Lifetime; 3.5% | $79,599/QALY | |||
DR screening every 3 years | DR screening every 5 years | Decision-analytic model | Lifetime; 3.5% | $45,574/QALY | |||
Nguyen et al., 2017/Singapore (83) | T2D and diabetic retinopathy | Telemedicine-based DR screening program, with real-time assessment of DR photographs by a centralized team supported by tele-ophthalmology IT infrastructure | Usual care (family physician assessment of DR) | Computer simulation model (decision tree and Markov) | Lifetime; 3% | Cost-saving | |
Scotland et al., 2016/Scotland (84) | T1D and T2D | Annual DR screening for those with no or mild retinopathy and biannual screening for observable retinopathy/maculopathy | Screening at 2-year intervals for those with no DR at two consecutive screening episodes | Markov computer simulation model (continuous-time hidden) | 30 years; 3.5% | $404,733/QALY | |
Screening at 2-year interval for those observed with no DR at two consecutive screening episodes | Screening at 2-year interval for those with no DR and no DR previously recorded | Markov computer simulation model (continuous-time hidden) | 30 years; 3.5% | $836,344/QALY | |||
Screening at 2-year interval for those with no DR and no DR previously recorded | Screening at 2-year interval for those with no DR | Markov computer simulation model (continuous-time hidden) | 30 years; 3.5% | $128,865/QALY | |||
van Katwyk et al., 2017/Canada (85) | Existing DM | DR screening by optometrists are publicly insured | Usual care (DR screening by primary care physician or referral to ophthalmologists are publicly insured) | Computer simulation probabilistic decision-analytic model | 30 years; 5% | $1,399/QALY | |
Foot ulcers | |||||||
Ragnarson Tennvall and Apelqvist, 2001/Sweden (86) | T1D and T2D, moderate to high risk (previous foot ulcer/amputation, neuropathy) | Optimal prevention of foot ulcer including foot inspection, appropriate footwear, treatment, and education | Usual care | Clinical and epidemiological data | 5 years; 0% | Cost-saving | |
Low risk (no specific risk factor) | >$127,000/QALY | ||||||
Ortegon et al., 2004/the Netherlands (87) | Newly diagnosed T2D + foot ulcer | Intensive glycemic control + optimal foot care | Standard care | Trial | Lifetime; 3% | $57,023/QALY | |
End-stage renal disease | |||||||
Borch-Johnsen et al., 1993/Germany (88) | T1D | Annual screening for microalbuminuria at 5 years after diabetes onset + ACEI treatment | Treatment of macroalbuminuria | Cohort | 30 years; 6% | Cost-saving | |
Kiberd and Jindal, 1995/Canada (89) | T1D | Screening for microalbuminuria + ACEI treatment | Treatment of hypertension and/or macroproteinuria | Clinical trial | Lifetime; 5% | $74,168/QALY | |
Golan et al., 1999/U.S. (90) | Newly diagnosed T2D | Treat patients with new diagnosis with ACEI | Screening for macroalbuminuria and treatment with ACEI | RCT | Lifetime; 3% | Cost-saving | |
Screening for microalbuminuria and treatment with ACEI | Screening for macroalbuminuria and treatment with ACEI | Cost-saving | |||||
Treat patients with new diagnosis with ACEI | Screening for microalbuminuria and treatment with ACEI | $13,843/QALY | |||||
Clark et al., 2000/Canada (91) | T1D | Province or territory paying for ACEI | Pay from out of pocket | Collaborative observational study using admin data base | 21 years; 5% | Cost-saving | |
Palmer et al., 2000/Switzerland (32) | T1D, high cholesterol, high systolic BP | Microalbuminuria monitoring, ACE treatment, conventional insulin therapy | Conventional insulin therapy | Literature review | Lifetime; 3% | Cost-saving | |
Palmer et al., 2003/Belgium, France (92) | T2D + macroalbuminuria + hypertension | Irbesartan | Standard therapy for hypertension | RCT | Lifetime; 3% | Cost-saving | |
Souchet et al., 2003/France (93) | T2D + nephropathy | Losartan | Placebo | Trial | 4 years; costs discounted 8%, benefits not discounted | Cost-saving | |
Dong et al., 2004/U.S. (94) | T1D | ACEI treatment starting at 1 year after diagnosis | Annual screening for microalbuminuria ACE treatment | Trial | Lifetime; 3% | $48,260/QALY, increased with lowering A1C level, at A1C level 9%, <$31,750/QALY | |
Palmer et al., 2004/U.K. (95) | T2D + hypertension + nephropathy | Irbesartan | Standard therapy for hypertension | RCT | 10 years; 6% for costs, 1.5% for benefits | Cost-saving | |
Palmer et al., 2004/U.S. (96) | T2D + hypertension + microalbuminuria | Irbesartan | Standard therapy for hypertension | RCT | 25 years; 3% | Cost-saving | |
Szucs et al., 2004/Switzerland (97) | T2D + nephropathy | Losartan | Placebo | Trial | 3.5 years; 0% | Cost-saving | |
Palmer et al., 2005/Spain (98) | T2D + microalbuminuria + hypertension | Irbesartan | Standard therapy for hypertension, no ACEI, AIIRA, or β-blockers | RCT | 25 years; 3% | Cost-saving | |
Rosen et al., 2005/U.S. (99) | Medicare population (T1D and T2D) | Medicare full payment for ACEI (target: ACEI use increased by at least 7.2%) | Pay from out of pocket | RCT | Lifetime; 3% | Cost-saving | |
Coyle et al., 2007/Canada (100) | T2D + hypertension + macronephropathy + micronephropathy | Irbesartan added at stage of microalbuminuria | Conventional treatment for diabetes and hypertension, no ACEI or AIIRAs | RCT | Lifetime; 5% | Cost-saving | |
Palmer et al., 2007/Hungary (101) | T2D + microalbuminuria | Adding irbesartan | Placebo + standard therapy for hypertension | RCT | 25 years; 5% | Cost-saving | |
Palmer et al., 2007/U.K. (102) | T2D + hypertension + microalbuminuria | Irbesartan | Standard therapy for hypertension | RCT | 25 years; 3.5% | Cost-saving | |
Irbesartan added at stage of overt nephropathy | Conventional treatment for diabetes and hypertension | Cost-saving | |||||
Irbesartan added at stage of microalbuminuria | Irbesartan added at stage of overt nephropathy | Cost-saving | |||||
Howard et al., 2010/Australia (41) | Individuals aged 50–69 years with T2D from the AusDiab study | Screening for proteinuria + addition of an ACEI | Usual care | Markov computer simulation model | Lifetime; 5% | $5,310/QALY | |
Comprehensive interventions | |||||||
Palmer et al., 2000/Switzerland (32) | T1D | Conventional glycemic control + ACEI therapy + eye screening and treatment | Conventional glycemic control | Lifetime; 3% | Cost-saving | ||
Intensive insulin therapy + ACEI therapy | Intensive insulin therapy | $59,055/LYG | |||||
Intensive insulin therapy + eye screening | Intensive insulin therapy | $64,262/LYG | |||||
Intensive insulin therapy + ACEI therapy + eye screening | Intensive insulin therapy | $63,246/LYG | |||||
Gozzoli et al., 2001/Switzerland (62) | T2D | Added education program, nephropathy screening, and ACEI therapy to standard antidiabetic care | Standard antidiabetic care | Lifetime; 0%, 3% | Cost-saving | ||
Added education program, nephropathy screening, ACEI therapy, and retinopathy screening and laser therapy to standard antidiabetic care | Standard antidiabetic care | Cost-saving | |||||
Multifactorial intervention included educational program, screening for nephropathy and retinopathy, control of CVD risk factors, early diagnosis and treatment of complications, and health education | Standard antidiabetic care | Cost-saving | |||||
Gaede et al., 2008/Denmark (103) | T2D and microalbuminuria (mean age 55 years) | Intensive treatment for 7.8 years (stepwise implementation of behavior modification and pharmacologic therapy targeting hyperglycemia, hypertension, dyslipidemia, and microalbuminuria, and 2° prevention of CVD with aspirin) | Standard care | Markov computer simulation model | Lifetime; 3% | $4,629/QALY $7,162/LYG | |
Tasosa et al., 2010/U.S. (104) | Newly diagnosed T2D, African American adults | Aggressive hypertension control with ACEI or β-blocker, glycemic control with insulin or sulfonylurea, hyperlipidemia treatment based on pravastatin and four physician visits with blood/lipid/biochemical profiles | Usual care | Markov computer simulation model | Lifetime; 3% | $33,912/QALY | |
Newly diagnosed T2D | Aggressive hypertension control with ACEI or β-blocker, glycemic control with insulin or sulfonylurea, hyperlipidemia treatment based on pravastatin and four physician visits with blood/lipid/biochemical profiles | Usual care | Markov computer simulation model | Lifetime; 3% | $51,587/QALY | ||
Giorda et al., 2014/Italy (105) | T2D | Physician-led 5-year quality-of-care scheme to improve A1C, BP, lipids, and BMI | Standard care | Computer simulation model | 50 years; 3% | Cost-saving | |
Laxy et al., 2017/U.K. (106) | Newly diagnosed T2D (mean age 61.5 years) from ADDITION-UK | Intensive lifestyle changes and medication adherence, delivered by a specialist team of doctors, nurses, dietitians (2 years) | Usual care | Trial/UKPDS Outcomes model | 10, 20, and 30 years; 3.5% | 10-year: $98,613/QALY 20-year: $39,378/QALY 30-year: $38,139/QALY | |
Integrated and coordinated care | |||||||
Coordinated care | |||||||
Mason et al., 2005/England (107) | T2D + hypertension | Policy to implement clinics led by specialist nurses to treat and control hypertension through consultation, medication review, condition assessment, and lifestyle advice | Usual care | RCT | Lifetime; 5% | $6,096/QALY | |
Diagnosed diabetes + dyslipidemia | Policy to implement clinics led by specialist nurses to treat and control hyperlipidemia by usual care | Usual care | $29,972/QALY | ||||
Gilmer et al., 2007/U.S. (108) | Diabetes, 48% Latinos, uninsured population | Culturally sensitive case management and self-management training program led by bilingual/bicultural medical assistant and registered dietitian stepped-care pharmacologic management of glucose and lipid levels and hypertension | Standard care | Cohort study | 40 years; 3% | $15,240/QALY | |
McRae et al., 2008/Australia (109) | T2D | Integrated care program whereby GPs serve as case manager and program facilitates case management via provision of info and education to GPs (5 years) | Usual care | Computer simulation model | 40 years; 5% | $9,058/LYG $10,871/QALE | |
Schouten et al., 2010/the Netherlands (110) | Existing T2D | Integrated diabetes care with teams of 5–6 providers that attended learning sessions in quality-improvement techniques and diabetes care, and access to endocrinologists and diabetes educators for patients unresponsive to treatment or with difficult-to-manage diabetes. | Usual care | Computer simulation model (Dutch diabetes model) | Lifetime; costs discounted at 4.5%, benefits discounted at 1.5% | Men: $11,806/QALY Women: $13,474/QALY | |
Kuo et al., 2011/U.S. (111) | T2D patients at U.S. Air Force base | Diabetes management using the Chronic Care Model for 3 years | Usual care | Markov computer simulation model | 20 years; 3% | $55,465/QALY | |
Haji et al., 2013/Australia (112) | T2D | High level of practice nurse involvement in T2D management in primary care setting | Low level of practice nurse involvement in T2D management in primary care setting | Computer simulation model (UKPDS Outcomes Model) | 40 years; 5% | Cost-saving | |
Slingerland et al., 2013/the Netherlands (113) | T2D + A1C <7% | Patient-centered medical care in which patients receive detailed “diabetes passports” based on national guidelines for 1 year | Usual care | Trial | Lifetime; costs discounted 3% | Intervention was associated with higher costs and fewer QALYs | |
T2D + A1C 7–8.5% | Patient-centered medical care in which patients receive detailed “diabetes passports” based on national guidelines for 1 year | Usual care | Trial | Lifetime | $23,764/QALY | ||
T2D + A1C >8.5% | Patient-centered medical care in which patients receive detailed “diabetes passports” based on national guidelines for 1 year | Usual care | Trial | Lifetime; costs discounted 3% | $7,622/QALY | ||
Yu et al., 2013/U.S. (114) | Existing T2D + A1C >7% | Addition of a pharmacist to patient's care (prescribed/adjusted medications, ordered laboratory work, ordered/administered immunizations, provided DM self-management education, and worked to optimize overall glycemic and cardiovascular care of patients) | Usual care (primary care physician only) | Markov computer simulation model | 10 years; costs discounted 3%, benefits discounted 5% | Cost-saving | |
Tsiachristas et al., 2014/the Netherlands (115) | T2D and Charlson comorbidity index 2.22 | DM management program consisting of personal coaching and motivational interviewing | DM management program consisting of lifestyle interventions, periodic discussion sessions between providers and patients | Program evaluation | Not reported | Cost-saving | |
Wilson et al., 2014/U.K. (116) | T2D | Intermediate care clinics for diabetes, in which diabetes specialist nurses worked closely with hospital-based specialist teams and community services (podiatry and dietetic services) to manage patients until risk factor control was achieved (18 months max) | Usual care | Trial | 18 months; no discounting | $13,552/QALY | |
Tao et al., 2015/U.K. (117) | Adults with screen-detected T2D | Intensive DM care (more frequent provider contact, interactive audit and feedback sessions, theory-based education materials, dietitian referrals, group programs) | Usual care | Computer simulation model | 30 years; 3.5% | $70,649/QALY | |
Hirsch et al., 2017/U.S. (118) | T2D + complications (average of 8 comorbidities) | Obtaining care in an endocrinologist-pharmacist collaborative practice (3 personalized 60-min visits over 6 months) | Usual PCP visits | Program evaluation; Archimedes computer simulation model (VA Health System) | 2, 5, and 10 years; 3% | Cost-saving | |
Cobden et al., 2010/U.S. (119) | Medicare adults with T2D and preexisting complications | Injectable insulin (human or analog), without adherence | Oral medications (metformin +/− sulfonylurea or TZD) without adherence | Markov computer simulation model | Lifetime (35 years max); 3% | $15,251/QALY | |
Injectable insulin (human or analog insulin), with adjustments for adherence | Oral medications (metformin +/− sulfonylurea or TZD), with adjustments for adherence | Markov computer simulation model | Lifetime (35 years max); 3% | $20,476/QALY | |||
Decision support | |||||||
Cleveringa et al., 2010/the Netherlands (120) | T2D | Diabetes care protocol, consisting of a diabetes consultation hour run by a practice nurse, a CDSS diagnostic and treatment algorithm based on Dutch T2D guidelines, a recall system, and a feedback at both practice and patient level every 3 months | Usual care | Computer microsimulation model | Lifetime; costs discounted 4%, benefits discounted 1.5% | $73,253/QALY $19,360/LYG | |
O'Reilly et al., 2012/Canada (121) | T2D | Computerized decision support system linked to EMR, shared between patients and physicians | Usual care | Computer simulation model (Ontario Diabetes Economic Model) | 40 years; 5% | $190,417/QALY $185,831/LYG | |
Olvey, 2014/U.S. (122) | DM and hypertension or high cholesterol | Patients spoke by phone to a Medication Management Center pharmacist who discussed ACEI/ARB and statin guidelines, and potential addition of those treatments based on final recommendation by the patient's physician | Patients received a letter listing current prescription info and advising to discuss treatments with their physician | Computer simulation model (decision tree and Monte Carlo) | 5 years; costs discounted 5%, benefits discounted 2.5% | $5,710/5-year treatment success | |
Peer support | |||||||
Gillespie et al., 2012/Ireland (123) | T2D | Group-based peer support in addition to standardized diabetes care for 2 years | Standard care | Computer simulation model | Lifetime; 3.5% | Cost-saving | |
Treatment of diabetes-related complications | |||||||
Cardiovascular disease | |||||||
Hlatky et al., 2009/U.S., Canada, Brazil, Mexico, Czech Republic, Austria (124) | T2D and CHD | Prompt coronary revascularization combined with intensive medical management for 4 years | Intensive medical management, with coronary revascularization at a later date if clinically indicated | Trial | Lifetime; costs discounted 3% | Within trial: control dominant (Lifetime: $810/LYG) | |
CABG with intensive medical management | Intensive medical management, with coronary revascularization at a later date if clinically indicated | Trial | Lifetime; costs discounted 3% | Within trial: control dominant Lifetime: $63,401/LYG | |||
Patients taking metformin or rosiglitazone or both for 4 years | Patients on insulin or sulfonylurea or both | Trial | Lifetime; costs discounted 3% | Within trial: $395,245/QALY Lifetime: $70,146/LYG | |||
Eye complications | |||||||
Sharma et al., 2001/U.S. (125) | Diabetic retinopathy (HMO) | Immediate vitrectomy for management of vitreous hemorrhage secondary to diabetic retinopathy | Deferral of vitrectomy | DRVS | Lifetime; 6% | $3,683/QALY | |
Mitchell et al., 2012/U.K. (126) | Existing DM and DME | Ranibizumab monotherapy | Laser photocoagulation | Markov computer simulation model (RESTORE Study) | 15 years; 3.5% | $45,264/QALY | |
Ranibizumab combined with laser therapy | Laser photocoagulation | Markov computer simulation model (RESTORE Study) | 15 years; 3.5% | $68,017/QALY | |||
Hutton et al., 2017/U.S. (127) | DM and proliferative diabetic retinopathy, with and without DME | Ranibizumab (0.5 mg) | Laser photocoagulation | Trial | 2 years; no discounting | With DME: $56,752/QALY Without DME: $677,108/QALY | |
Foot ulcers | |||||||
Habacher et al., 2007/Austria (128) | Newly diagnosed diabetic food ulcer | Intensified treatment by international consensus on diabetic foot care | Standard treatment | Retrospective of patient records | 15 years; 0–8% | Cost-saving | |
O'Connor et al., 2008/U.S. (129) | DM and painful diabetic peripheral neuropathy | Duloxetine 60 mg 1×/day | Desipramine 100 mg 1×/day | Computer simulation model (decision tree) | 3 months; no discounting | $67,188/QALY | |
Pregabalin 100 mg 1×/day | Desipramine 100 mg 1×/day | Computer simulation model (decision tree) | 3 months; no discounting | Intervention associated with higher cost, worse outcome | |||
Gabapentin 800 mg 1×/day | Desipramine 100 mg 1×/day | Computer simulation model (decision tree) | 3 months; no discounting | Intervention associated with higher cost, worse outcome | |||
Cheng et al., 2017/Australia (130) | Simulated cohort of existing DM and at high risk of developing foot ulcers | Optimal care for foot ulcers and patient education | Usual care | Markov computer simulation model | 5 years; 5% | Cost-saving | |
Addressing diabetes comorbidities | |||||||
Obesity | |||||||
Anselmino et al., 2009/Austria, Italy, Spain (131) | T2D and BMI >35 kg/m2 | Gastric banding surgery | Usual care | Computer simulation model (deterministic linear algorithm) | 5 years; 3.5% | Austria: (−$5,027)/QALY, cost-saving Italy: (−$1,945)/QALY cost-saving Spain: $2,558/QALY | |
Gastric bypass surgery | Usual care | Computer simulation model (deterministic linear algorithm) | 5 years; 3.5% | Austria: (−$2,542)/QALY, cost-saving Italy: (−$2,189)/QALY, cost-saving Spain: $4,680/QALY | |||
Ikramuddin et al., 2009/U.S. (132) | T2D and obesity | Gastric bypass surgery | Standard medical management | Computer simulation model (CORE Diabetes Model) | 35 years; 3% | $29,641/QALY $40,032/LYG | |
Keating et al., 2009/Australia (133) | T2D and obesity (class I and II) | Gastric band surgery + conventional therapy for 2 years | Conventional therapy for 2 years | Computer simulation model | Lifetime; 3% | Cost-saving | |
Hoerger et al., 2010/U.S. (134) | Newly diagnosed or existing T2D and BMI ≥35 kg/m2 | Gastric bypass/gastric banding surgery | Standard care | Computer simulation model | Lifetime; 3% | For newly diagnosed DM: $10,254/QALY for gastric bypass $16,115/QALY for gastric banding For existing DM: $17,580/QALY for gastric bypass $19,045/QALY for gastric banding | |
Pollock et al., 2013/U.K. (135) | T2D and obesity | Gastric banding surgery | Standard care | Computer simulation model (CORE Diabetes Model) | 40 years; 3.5% | $6,785/QALY | |
Borisenko et al., 2015/Sweden (136) | T2D and obesity | Bariatric surgery | No surgery | Decision-analytic model using Markov processes | Lifetime; 3% | Bariatric surgery becomes cost-effective after 2 years ($39,604/QALY) and cost-saving after 17 years | |
James et al., 2017/Australia (137) | Simulated cohort of 30-year-old Australian females with T2D and obesity | Gastric banding surgery | Usual care (pharmacotherapy, diet, exercise management) | Markov computer simulation model | Lifetime; 5% | Cost-saving | |
Gastric bypass surgery | Usual care (pharmacotherapy, diet, exercise management) | Markov computer simulation model | Lifetime; 5% | Cost-saving | |||
Sleeve gastrectomy surgery | Usual care (pharmacotherapy, diet, exercise management) | Markov computer simulation model | Lifetime; 5% | Cost-saving | |||
Wentworth et al., 2017/U.S. (138) | T2D and overweight | Gastric banding surgery | Usual care | Computer simulation model (UKPDS Outcomes Model) | 2 and 10 years; 3% | Within 2-year trial: $100,050/QALY 5-year simulation: $55,120/QALY 10-year simulation: $30,747/QALY 15-year simulation: $23,320/QALY | |
Mental health | |||||||
Katon et al., 2006/U.S. (139) | Depression + poorly controlled DM or CHD | Multicondition collaborative treatment program led by a physician-supervised registered nurse and including patient education to promote self-care for 2 years (TEAMCare) | Usual care | Trial | NA | Cost-saving | |
Johnson et al., 2016/Canada (140) | T2D + depressive symptoms (PHQ ≥10) | Screening for depression + enhanced care (follow-up with family physician) | Usual care | Trial | 1 year; no discounting | $91,270/QALY | |
Screening for depression + coordinated, collaborative care led by a nurse care manager, in consultation with psychiatrists/endocrinologists (adapted TEAMCare) | Usual care | Trial | 1 year; no discounting | $29,160/QALY | |||
Screening for depression + coordinated, collaborative care led by a nurse care manager, in consultation with psychiatrists/endocrinologists (adapted TEAMCare) | Screening for depression + enhanced care (follow-up with family physician) | Trial | 1 year; no discounting | $18,980/QALY | |||
Kearns et al., 2017/U.K. (141) | Simulated cohort of existing T2D | Collaborative care | Usual care | Computer simulation (discrete event) | Lifetime; 3.5% | $18,814/QALY | |
Improved opportunistic screening for depression | Usual care | Computer simulation (discrete event) | Lifetime; 3.5% | $111,180/QALY | |||
Collaborative care + improved opportunistic screening for depression | Usual care | Computer simulation (discrete event) | Lifetime; 3.5% | $65,201/QALY | |||
Sleep apnea | |||||||
Guest et al., 2014/U.K. (142) | T2D with obstructive sleep apnea | Treatment with CPAP for 5 years | Standard care | Program evaluation/trial | 5 years; no discounting | $27,750/QALY |
A1C, hemoglobin A1c test; AIIRA, angiotensin II receptor antagonist; AusDiab, Australian Diabetes, Obesity, and Lifestyle Study; BP, blood pressure; CABG, coronary artery bypass graft; CDSS, clinical decision support system; CHD, coronary heart disease; DCCT, Diabetes Control and Complications Trial; DRVS, Diabetic Retinopathy Vitrectomy Study; ECG, electrocardiogram; EMR, electronic medical record; FPG, fasting plasma glucose; DM, diabetes; DME, diabetic macular edema; DPP, Diabetes Prevention Program; DPP-4, dipeptidyl peptidase 4; DR, diabetic retinopathy; GCT, glucose challenge test; GDM, gestational diabetes mellitus; GP, general practitioner; GTT, glucose tolerance test; HMO, health maintenance organization; IADPSB, International Association of the Diabetes and Pregnancy Study Groups; ISO, International Organization for Standardization; IT, information technology; LEADER, Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results; max, maximum; OGTT, oral glucose tolerance test; PHQ, patient health questionnaire; preDM, prediabetes; QALE, quality-adjusted life expectancy; RCT, randomized controlled trial; TZD, thiazolidinedione; UKPDS, UK Prospective Diabetes Study.
Supplementary Fig. 1 provides a summary of the numbers of studies across the four broad intervention categories—screening for undiagnosed diabetes (8 studies), managing diabetes and risk factors to prevent diabetes-related complications (71 studies), screening for and early treatment of diabetes complications (33 studies), and treating diabetes-related complications and comorbidities (19 studies)—as well as how the number of articles in each category changed from the previous review (1985–2008) to the current review (1985–2017). Except for smoking cessation, the number of articles in every category increased over time, and five new categories emerged: preventing CVD complications, treating CVD complications, and addressing comorbidities of obesity, mental health, and sleep apnea.
In Table 2, we classified each of the ADA-recommended interventions based on their levels of CE and strength of evidence by intervention category, using all studies over the period 1985–2017. To facilitate use by clinicians and decision makers, we describe the findings across each of the four intervention categories from a health system perspective.
Summary of the CE studies by intervention (U.S. and non-U.S. high-income country studies)
Intervention . | Comparison . | Intervention population . | Level of recommendation by ADA . | Range of the CE ratios . | Median of the CE ratios (no. of studies) . | CE based on previous review (no. of studies) . |
---|---|---|---|---|---|---|
Strong evidence | ||||||
Cost-saving | ||||||
ACEI/ARB therapy for intensive hypertension control | Standard hypertension control | T2D with hypertension | A | Cost-saving–$254/QALY | Cost-saving (6) | Cost-saving (4) |
ACEI/ARB therapy to prevent CKD and/or ESRD | No ACEI/ARB therapy | T2D | A | Cost-saving–$5,310/QALY | Cost-saving (11) | Cost-saving (11) |
Comprehensive foot care and patient education to prevent and treat foot ulcers | Usual care | DM and moderate/high risk of developing foot ulcers | B | Cost-saving | Cost-saving (3) | Cost-saving (2) |
Telemedicine for DR screening | Office screening for DR | DM | B | Cost-saving–$7,781/QALY | Cost-saving (4) | New finding (previously only supportive evidence) |
Bariatric surgery | No bariatric surgery | T2D and obesity | A | Cost-saving–$29,641/QALY | Cost-saving (7) | New finding |
Very cost-effective | ||||||
Intensive glycemic control | Conventional glycemic control | T2D, newly diagnosed and young age | A | Cost-saving–$78,740 | $4,318/QALY (6) | $4,318/QALY (6) |
Multicomponent interventions (behavior change/education and pharma therapy targeting hyperglycemia, hypertension, dyslipidemia, microalbuminuria, nephropathy/retinopathy, 2° prevention of CVD with aspirin) | Usual care | T1D/T2D | A: behavior modification A: aspirin A: risk factor control B: pharma therapy B: education | Cost-saving–$58,587/QALY | $2,315/QALY (6) | Cost-saving (2) |
Statin therapy | No statin therapy | T2D with hyperlipidemia + CVD history | A | Cost-saving–$15,621/LYG | $4,627/LYG (4) | Very cost-effective (3) |
Diabetes self-management education | Usual care | T1D/T2D | B | Cost-saving–$55,726/QALY | $5,047/QALY (11) | Cost-saving, supportive evidence (2) |
T2D screening every 3 years starting at age 45 years (as recommended by ADA) | No screening | U.S. population without DM | B | $2,088–$13,707/QALY | $7,898/QALY (2) | Very cost-effective (1) |
T2D screening every 1 year starting at age 45 years | T2D screening every 3 years starting at age 45 years | U.S. population without DM | B | $8,139/QALY | $8,139/QALY (2) | Very cost-effective (1) |
Screening for eye complications every 1–2 years (as recommended by ADA) | No screening | T1D/T2D | B | Cost-saving–$50,165/QALY | $8,763/QALY (7) | Very cost-effective (5) |
Integrated, patient-centered care (high level of nurse/pharmacist involvement) and based on Chronic Care Model | Usual care | T2D | B | Cost-saving–$55,465/QALY | $11,339/QALY (8) | Very cost-effective, supportive evidence |
Smoking cessation | No smoking cessation | T2D | A, B | <$31,750–$114,046/QALY | <$31,750/QALY(1) | Very cost-effective (1) |
Daily aspirin use as primary prevention for cardiovascular complications | Usual care | T2D | B | $2,395/QALY | $2,395/QALY (1) | New finding |
SMBG 3×/day | SMBG 1×/day | T2D adults | B | Cost-saving–$9,201/QALY | $3,719 (4) | New finding |
Intensive glycemic control | Conventional insulin therapy | T2D aged ≥50 years | A | $15,398/QALY | $15,398/QALY (1) | New finding |
Collaborative care for depression (TEAMCare) | Usual care | T2D + depression | B | Cost-saving–$29,160/QALY | $18,814/QALY (3) | New finding |
Cost-effective | ||||||
Intensive glycemic control | Usual care | T1D | A | $12,954–$64,516/QALY | $41,339/QALY (4) | Cost-effective (4) |
Marginally cost-effective | ||||||
Statin therapy | No statin therapy | T2D with hyperlipidemia without CVD history | A | $4,445–$98,906/QALY | $67,873/QALY (3) | Cost-effective (3) |
Ranibizumab treatment | Panretinal photocoagulation | DM and DR, with DME | A | $45,264–$56,752/QALY | $51,008/QALY (2) | New finding |
Duloxetine for the treatment of PDPN | Desipramine | DM and PDPN | A | $67,188/QALY | $67,188/QALY (1) | New finding |
Not cost-effective | ||||||
Universal opportunistic screening for undiagnosed T2D | Targeted screening in persons with hypertension | U.S. population ≥45 years | B | $89,535–$888,746/QALY | >$100,000/QALY (1) | Not cost-effective (1) |
Universal opportunistic screening for undiagnosed T2D and ensuing treatment | No screening | U.S. population ≥45 years | B | $89,027–$1,178,560/QALY | >$100,000/QALY (2) | Not cost-effective (2) |
Supportive evidence | ||||||
Cost-saving | ||||||
Reimbursement for ACEI by public insurance | Paying out of pocket | T1D/T2D | B/E | Cost-saving | Cost-saving (1) | Cost-saving (1) |
Group-based peer support (9 group meetings led by peer supporters in general practice) | Usual care | T2D | B | Cost-saving | Cost-saving (1) | Supportive—recommendation updated in new ADA |
Very cost-effective | ||||||
Statin treatment at T2D diagnosis (as primary prevention) | No lipid-regulating treatment | T2D adults | No recommendation level | $3,294/QALY; $3,640–$84,012/QALY for different age-groups | $3,294/QALY (1) | New finding |
Bariatric surgery – gastric bypass/gastric banding | No surgery | T2D and overweight | A | $23,320/QALY | $23,320/QALY (1) | New finding |
Cost-effective | ||||||
CPAP for the treatment of obstructive sleep apnea | Usual care | T2D with obstructive sleep apnea | E | $27,750/QALY | $27,750/QALY (1) | New finding |
Marginally cost-effective | ||||||
Adhering to National Cholesterol Education Program Adult Treatment Panel III guidelines | Usual care | T2D adults with mixed dyslipidemia | A | $67,873/QALY | $67,873/QALY (1) | New finding |
Uncertain evidence | ||||||
Screening for GDM | No screening | 30-year-old pregnant women between 24–28 weeks | C | Cost-saving–$19,746/QALY | Cost-saving (6) | Cost-saving (5) |
SMBG 1×/day, provision of device and 1.29 strips per day | Usual care | T1D/T2D patients not using insulin | B: SMBG E: providing device and strips | $77,684–$130,820/QALY | >$100,000/QALY (2) | New finding |
Computerized decision support system linked to EHR, shared between patients and physicians | Usual care | T2D | B | $190,417/QALY | $190,417/QALY (1) | New finding |
Intervention . | Comparison . | Intervention population . | Level of recommendation by ADA . | Range of the CE ratios . | Median of the CE ratios (no. of studies) . | CE based on previous review (no. of studies) . |
---|---|---|---|---|---|---|
Strong evidence | ||||||
Cost-saving | ||||||
ACEI/ARB therapy for intensive hypertension control | Standard hypertension control | T2D with hypertension | A | Cost-saving–$254/QALY | Cost-saving (6) | Cost-saving (4) |
ACEI/ARB therapy to prevent CKD and/or ESRD | No ACEI/ARB therapy | T2D | A | Cost-saving–$5,310/QALY | Cost-saving (11) | Cost-saving (11) |
Comprehensive foot care and patient education to prevent and treat foot ulcers | Usual care | DM and moderate/high risk of developing foot ulcers | B | Cost-saving | Cost-saving (3) | Cost-saving (2) |
Telemedicine for DR screening | Office screening for DR | DM | B | Cost-saving–$7,781/QALY | Cost-saving (4) | New finding (previously only supportive evidence) |
Bariatric surgery | No bariatric surgery | T2D and obesity | A | Cost-saving–$29,641/QALY | Cost-saving (7) | New finding |
Very cost-effective | ||||||
Intensive glycemic control | Conventional glycemic control | T2D, newly diagnosed and young age | A | Cost-saving–$78,740 | $4,318/QALY (6) | $4,318/QALY (6) |
Multicomponent interventions (behavior change/education and pharma therapy targeting hyperglycemia, hypertension, dyslipidemia, microalbuminuria, nephropathy/retinopathy, 2° prevention of CVD with aspirin) | Usual care | T1D/T2D | A: behavior modification A: aspirin A: risk factor control B: pharma therapy B: education | Cost-saving–$58,587/QALY | $2,315/QALY (6) | Cost-saving (2) |
Statin therapy | No statin therapy | T2D with hyperlipidemia + CVD history | A | Cost-saving–$15,621/LYG | $4,627/LYG (4) | Very cost-effective (3) |
Diabetes self-management education | Usual care | T1D/T2D | B | Cost-saving–$55,726/QALY | $5,047/QALY (11) | Cost-saving, supportive evidence (2) |
T2D screening every 3 years starting at age 45 years (as recommended by ADA) | No screening | U.S. population without DM | B | $2,088–$13,707/QALY | $7,898/QALY (2) | Very cost-effective (1) |
T2D screening every 1 year starting at age 45 years | T2D screening every 3 years starting at age 45 years | U.S. population without DM | B | $8,139/QALY | $8,139/QALY (2) | Very cost-effective (1) |
Screening for eye complications every 1–2 years (as recommended by ADA) | No screening | T1D/T2D | B | Cost-saving–$50,165/QALY | $8,763/QALY (7) | Very cost-effective (5) |
Integrated, patient-centered care (high level of nurse/pharmacist involvement) and based on Chronic Care Model | Usual care | T2D | B | Cost-saving–$55,465/QALY | $11,339/QALY (8) | Very cost-effective, supportive evidence |
Smoking cessation | No smoking cessation | T2D | A, B | <$31,750–$114,046/QALY | <$31,750/QALY(1) | Very cost-effective (1) |
Daily aspirin use as primary prevention for cardiovascular complications | Usual care | T2D | B | $2,395/QALY | $2,395/QALY (1) | New finding |
SMBG 3×/day | SMBG 1×/day | T2D adults | B | Cost-saving–$9,201/QALY | $3,719 (4) | New finding |
Intensive glycemic control | Conventional insulin therapy | T2D aged ≥50 years | A | $15,398/QALY | $15,398/QALY (1) | New finding |
Collaborative care for depression (TEAMCare) | Usual care | T2D + depression | B | Cost-saving–$29,160/QALY | $18,814/QALY (3) | New finding |
Cost-effective | ||||||
Intensive glycemic control | Usual care | T1D | A | $12,954–$64,516/QALY | $41,339/QALY (4) | Cost-effective (4) |
Marginally cost-effective | ||||||
Statin therapy | No statin therapy | T2D with hyperlipidemia without CVD history | A | $4,445–$98,906/QALY | $67,873/QALY (3) | Cost-effective (3) |
Ranibizumab treatment | Panretinal photocoagulation | DM and DR, with DME | A | $45,264–$56,752/QALY | $51,008/QALY (2) | New finding |
Duloxetine for the treatment of PDPN | Desipramine | DM and PDPN | A | $67,188/QALY | $67,188/QALY (1) | New finding |
Not cost-effective | ||||||
Universal opportunistic screening for undiagnosed T2D | Targeted screening in persons with hypertension | U.S. population ≥45 years | B | $89,535–$888,746/QALY | >$100,000/QALY (1) | Not cost-effective (1) |
Universal opportunistic screening for undiagnosed T2D and ensuing treatment | No screening | U.S. population ≥45 years | B | $89,027–$1,178,560/QALY | >$100,000/QALY (2) | Not cost-effective (2) |
Supportive evidence | ||||||
Cost-saving | ||||||
Reimbursement for ACEI by public insurance | Paying out of pocket | T1D/T2D | B/E | Cost-saving | Cost-saving (1) | Cost-saving (1) |
Group-based peer support (9 group meetings led by peer supporters in general practice) | Usual care | T2D | B | Cost-saving | Cost-saving (1) | Supportive—recommendation updated in new ADA |
Very cost-effective | ||||||
Statin treatment at T2D diagnosis (as primary prevention) | No lipid-regulating treatment | T2D adults | No recommendation level | $3,294/QALY; $3,640–$84,012/QALY for different age-groups | $3,294/QALY (1) | New finding |
Bariatric surgery – gastric bypass/gastric banding | No surgery | T2D and overweight | A | $23,320/QALY | $23,320/QALY (1) | New finding |
Cost-effective | ||||||
CPAP for the treatment of obstructive sleep apnea | Usual care | T2D with obstructive sleep apnea | E | $27,750/QALY | $27,750/QALY (1) | New finding |
Marginally cost-effective | ||||||
Adhering to National Cholesterol Education Program Adult Treatment Panel III guidelines | Usual care | T2D adults with mixed dyslipidemia | A | $67,873/QALY | $67,873/QALY (1) | New finding |
Uncertain evidence | ||||||
Screening for GDM | No screening | 30-year-old pregnant women between 24–28 weeks | C | Cost-saving–$19,746/QALY | Cost-saving (6) | Cost-saving (5) |
SMBG 1×/day, provision of device and 1.29 strips per day | Usual care | T1D/T2D patients not using insulin | B: SMBG E: providing device and strips | $77,684–$130,820/QALY | >$100,000/QALY (2) | New finding |
Computerized decision support system linked to EHR, shared between patients and physicians | Usual care | T2D | B | $190,417/QALY | $190,417/QALY (1) | New finding |
ADA, American Diabetes Association Standards of Care 2018; DM, diabetes; DME, diabetic macular edema; DR, diabetic retinopathy; EHR, electronic health record; PDPN, painful diabetic peripheral neuropathy. Cost-saving is defined as an intervention that generates a better health outcome and costs less than the comparison intervention or is cost neutral (ICER = 0); very cost-effective, 0 < ICER ≤ $25,000 per QALY or LYG; cost-effective, $25,000 < ICER ≤ $50,000 per QALY or LYG; marginally cost-effective, $50,000 < ICER ≤ $100,000 per QALY or LYG; or not cost-effective, >$100,000 per QALY or LYG. A, as defined in Standards of Care 2018: clear evidence from well-conducted, generalizable, randomized controlled trials that are adequately powered; compelling nonexperimental evidence, i.e., “all or none” rule developed by the Centre for Evidence-Based Medicine at Oxford; supportive evidence from well-conducted randomized controlled trials that are adequately powered. B, as defined in Standards of Care 2018: supportive evidence from well-conducted cohort studies; supportive evidence from a well-conducted case-control study. C, as defined in Standards Care 2018: supportive evidence from poorly controlled or uncontrolled studies; conflicting evidence with the weight of evidence supporting the recommendation. E, as defined in Standards of Care 2018: expert consensus or clinical experience.
Strong Evidence
Screening for Undiagnosed Diabetes
Screening for T2D every 3 years starting at age 45 years for the U.S. population without diabetes, compared with no screening, had strong evidence of being very cost-effective at $7,898/QALY (every 1 year compared with every 3 years was also very cost-effective at $8,139/QALY). On the other hand, there was strong evidence that universal opportunistic screening for undiagnosed T2D among the U.S. population (whether or not followed by treatment), compared with targeted screening in high-risk individuals, was not cost-effective (>$100,000/QALY).
Managing Diabetes and Risk Factors to Prevent Diabetes-Related Complications
For interventions to manage diabetes and risk factors to prevent complications, the evidence was mixed. We found strong evidence that DSME for individuals with diabetes, compared with usual care, is very cost-effective ($5,047/QALY). Additionally, there were several new studies on the daily frequency of SMBG, which led to the new finding that SMBG three times per day, compared with SMBG once per day, is very cost-effective ($3,719/QALY) among adults with T2D currently taking insulin.
The ICERs for intensively managing glycemia varied according to a patient’s age, duration of diabetes, and diabetes type (1 or 2). We found that intensive glycemic management compared with conventional management was very cost-effective among young individuals with newly diagnosed T2D ($4,318/QALY) and older individuals (aged ≥50 years) with T2D ($15,398/QALY) and was cost-effective when given to individuals with T1D regardless of age ($41,339/QALY).
For blood pressure management, ACE inhibitor (ACEI) and angiotensin receptor blocker (ARB) therapies, used either for intensive hypertension management (compared with suboptimal blood pressure management) or to prevent chronic kidney disease and/or ESRD in patients with albuminuria, compared with no ACEI/ARB therapy, emerged with strong evidence of being cost-saving.
Over the period 1985–2017, studies comparing multicomponent interventions (including behavior change and medication adherence to improve glycemia, blood pressure/CVD, lipids, and nephropathy/retinopathy prevention and screening together) with usual care have shown a range of value achieved from cost-effective to cost-saving. Overall, however, we found that these multicomponent interventions were on average very cost-effective ($2,315/QALY) for individuals with T1D and T2D compared with usual care.
Diabetes management interventions that remained consistent with the previous review as very cost-effective were 1) integrated, patient-centered care based on the Chronic Care Model for individuals with T2D compared with usual care ($11,339/QALY), and 2) smoking cessation for individuals with diabetes compared with no smoking cessation (<$31,750/QALY).
Screening for and Early Treatment of Diabetes-Related Complications
We found strong evidence for two cost-saving screening and early treatment interventions: 1) comprehensive foot care and patient education to prevent and treat foot ulcers among individuals with diabetes and at moderate/high risk of developing foot ulcers, compared with usual care, and 2) telemedicine for diabetic retinopathy screening among individuals with diabetes compared with office screening for diabetic retinopathy. Additionally, we found strong evidence that screening for eye complications every 1–2 years for individuals with diabetes, compared with no screening, is very cost-effective ($8,763/QALY).
Treatment of Diabetes-Related Complications and Comorbidities
Statin therapy for secondary prevention of CVD—i.e., in individuals with T2D and a history of CVD—compared with no statin therapy remained consistent from the previous review as very cost-effective ($4,627/QALY), while among individuals with T2D, hyperlipidemia, and no history of CVD, statin therapy was marginally cost-effective ($67,873/QALY). A new finding in this review was that daily aspirin use for primary prevention of CVD among individuals with T2D, compared with usual care, was very cost-effective ($2,395/QALY).
There were eight studies evaluating the CE of bariatric surgery in individuals with T2D and obesity (BMI ≥30 kg/m2) compared with no bariatric surgery. All eight studies found this intervention to be cost-saving. Additionally, three studies evaluated the CE of collaborative care models to comanage depression in individuals with T2D and depression compared with usual care and found such treatment to be very cost-effective ($18,814/QALY).
Studies on two new ADA-recommended drugs (ranibizumab for diabetic retinopathy and duloxetine for painful diabetic peripheral neuropathy) were also included in this review. Both—ranibizumab compared with panretinal photocoagulation and duloxetine compared with desipramine—were found to be marginally cost-effective ($51,008/QALY and $67,188/QALY, respectively).
Supportive and Uncertain Evidence
There were nine specific interventions for which the level of CE was based on supportive evidence. Among these, cost-saving interventions are 1) reimbursement for ACEI by public insurance for individuals with diabetes compared with paying out of pocket and 2) group-based peer support for individuals with T2D compared with usual care. Very cost-effective interventions, based on supportive evidence, were both new findings in this updated review: 1) statin treatment at T2D diagnosis compared with no lipid-regulating treatment ($3,294/QALY) and 2) bariatric surgery for individuals with T2D and overweight compared with no surgery ($23,320/QALY). Continuous positive airway pressure (CPAP) therapy for individuals with T2D and obstructive sleep apnea compared with usual care was also cost-effective ($27,750/QALY). Adhering to the National Cholesterol Education Program Adult Treatment Panel III guidelines for adults with T2D and mixed dyslipidemia compared with usual care was marginally cost-effective ($67,873/QALY).
There were three specific interventions in the “uncertain evidence” category: 1) screening a 30-year-old pregnant woman between 24–28 weeks’ gestation (base case) for gestational diabetes mellitus compared with no screening (cost-saving); 2) SMBG once per day and provision of monitoring devices and strips for individuals with T1D and for those with T2D not using insulin compared with usual care (SMBG once per day without provision of devices and strips), at >$100,000/QALY; and 3) computerized decision-support systems linked to electronic health records and shared between patients and physicians ($190,417/QALY).
Conclusions
Our systematic review provides an updated understanding of the potential value of interventions to manage and treat diabetes from a health system perspective. Since the last review in 2010, the evidence that interventions to manage diabetes are cost-effective has grown in terms of additional evaluations to bolster existing evidence, as well as new economic evaluations of novel interventions and methods of care delivery. ACEI/ARB therapy compared with standard hypertension management, comprehensive foot care compared with usual care, and intensive glycemic management compared with conventional therapy are confirmed as very cost-effective interventions, while multicomponent interventions compared with usual care, statin therapy for secondary prevention compared with no statin therapy, T2D screening every 3 years compared with no screening, and screening for eye complications compared with no screening are confirmed as very cost-effective interventions. New findings include telemedicine for diabetic retinopathy screening and bariatric surgery for type 2 diabetes and BMI ≥30 kg/m2 as cost-saving interventions and aspirin use for primary prevention of cardiovascular complications, SMBG three times per day for insulin-treated patients (compared with once per day), intensive glycemic management among those aged ≥50 years, and collaborative care for depression as very cost-effective interventions. This review complements professional treatment recommendations and can assist clinicians and payers in prioritizing interventions in an evidence-based manner that may lead to better allocation of health care resources.
Figure 2 is a comprehensive guide to our findings. Overall, the ADA-recommended interventions included in the previously published review remain cost-saving, very cost-effective, or cost-effective. The strength of the evidence improved from supportive to strong for both DSME (compared with usual care) and integrated, patient-centered care based on the Chronic Care Model (compared with usual care) due to additional studies on these topics during the 2008–2017 time period. Interventions that are cost-saving should be implemented, and those that are very cost-effective or cost-effective based on strong evidence warrant consideration for implementation. ADA-recommended interventions rated as cost-saving, very cost-effective, or cost-effective with supportive evidence should be adopted if extra resources are available or if similar interventions with strong evidence are unavailable or infeasible in a specific setting.
Summary of the CE of interventions (strong evidence only). CKD, chronic kidney disease; DM, diabetes; DR, diabetic retinopathy; PDPN, painful diabetic peripheral neuropathy; undx, undiagnosed.
Summary of the CE of interventions (strong evidence only). CKD, chronic kidney disease; DM, diabetes; DR, diabetic retinopathy; PDPN, painful diabetic peripheral neuropathy; undx, undiagnosed.
Our review highlighted the value associated with new and innovative interventions to manage and treat diabetes, including technology-related innovations and those focused on addressing diabetes-related comorbidities. The focus on technological innovations and diabetes-related comorbidities is well-aligned with the ADA SOC 2019 (14) (as well as the recently updated ADA SOC 2020 [3,14]). In addition, the inclusion of collaborative care models for depression is a big step toward acknowledging and addressing the comorbid conditions of diabetes and depression, which is increasingly being seen as an important consideration in the care of people with diabetes (15).
Multicomponent interventions are also featured prominently in the current review and may have implications for delivery system design, especially in the context of persistent gaps in achievement of diabetes care goals (16,17). Among individuals with diabetes, interventions that included a combination of practice change, behavior change and education, pharmacologic therapy targeting hyperglycemia, hypertension, dyslipidemia, microalbuminuria, or nephropathy/retinopathy, and secondary prevention of CVD with aspirin were very cost-effective compared with usual care, based on strong evidence. Of note, our review predates the ASCEND (A Study of Cardiovascular Events in Diabetes) trial, which showed that the primary CVD prevention benefits of aspirin were offset by the increased risk of major bleeding events (18).
In some cases, CE evaluations may help to provide more insight into how ADA-recommended interventions might be prioritized for specific populations receiving the intervention. For example, bariatric surgery was cost-saving among individuals with T2D and obesity but only very cost-effective among individuals with T2D and overweight, likely due to larger health risks posed by obesity. We also noted differences in the CE of statin therapy for individuals with diabetes with and without CVD; when used as secondary prevention, there is clear value to statin use (cost-saving). However, for primary prevention, statin use has been found to be less cost-effective. We found that the CE of intensive glycemic management (with a goal of reducing A1C values) depends on age and duration since diabetes diagnosis. Among young individuals with newly diagnosed T2D, intensive glycemic management, compared with conventional insulin therapy, is cost-saving; indeed, recent research shows that earlier intensive management is associated with lower long-term risk of complications (19). For older individuals (aged ≥50 years) with T2D and shorter life expectancy, the ability to see benefits of intensive glycemic management is limited, in part because cardiovascular or mortality benefits may not be seen for at least 10 years (20). In individuals with T1D, intensive glycemic management compared with conventional insulin therapy remains cost-effective.
There are a few key areas that future economic evaluations of diabetes should consider. First, more studies are needed to evaluate the CE of interventions that fell in the “supportive” or “uncertain” evidence categories. In cases where interventions have uncertain value due to a small number of studies (i.e., incomplete knowledge), adding to the evidence base could help to clarify their value. There are also a number of new, efficacious medications and treatments for the management of glycemia (glucagon-like peptide 1 receptor agonists, sodium–glucose cotransporter 2 inhibitors), lipids (PSK9 inhibitors), and heart failure (sacubitril/valsartan [Entresto]) that were not included in this review because there are no published CE studies. For studies with weaker efficacy data, further efficacy studies are needed.
Second, more studies are needed that address interventions in real-world settings, as our current review is predominantly based on randomized controlled trials or computer simulation models. In real life, however, there are many factors to consider in addition to the CE of an intervention, such as 1) coverage for the intervention in question (determine access), 2) motivation (side effects or mode of delivery [e.g., injectable versus oral] may deter patients from taking specific medications), and 3) whether the risk reduction in real life is similar to what was observed in trials and models (i.e., effectiveness versus efficacy). These are many of the unknowns that clinicians and policy makers must consider as they attempt to use the data from this review in practice.
Third, there may be additional cost-effective interventions that exist but have not been studied or about which the right questions are not being asked. For example, the 2010 review included one article regarding the CE of smoking cessation, which was found to be very cost-effective. There were no additional articles in this category from 2008–2017, likely because it is universally understood that smoking cessation is good and thus no one would be compelled to argue its value; a more interesting contemporary economic question might be to inquire how often a smoking cessation intervention should be implemented for it to be most cost-effective and most adoptable by clinicians and their patients.
The CE of an intervention in decision- making is important, but it is not the only factor to consider. CE analysis does not address equity in the distribution of costs and the benefits of an intervention, societal or personal willingness to pay, social and legal aspects, or ethical issues associated with each intervention. However, with an eye toward finding diabetes management and treatment interventions that can best increase the value of our health care dollars, this review of the most up-to-date available evidence can help to guide clinicians and policy makers toward the most cost-effective use of their prescriptions and health care dollars.
The findings and conclusions are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
See accompanying article, p. 1593.
This article contains supplementary material online at https://doi.org/10.2337/figshare.12081801.
Article Information
Acknowledgments. This work is a collaboration between the Centers for Disease Control and Prevention and the ADA. The authors thank the external and internal reviewers for their valuable comments during the review process. The authors thank Rui Li for generously sharing materials from her previous review and William Thomas for his timely help with the literature search (both from Centers for Disease Control and Prevention).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. K.R.S., X.Zho., B.P.N., S.J., K.P., and X.Zha. reviewed the abstracts and full text of all articles for inclusion and abstracted the data. X.Zho. and B.P.N. performed the literature search and rated the quality of the studies for inclusion. K.R.S., M.K.A., X.Zho., E.W.G., A.L.A., and P.Z. interpreted the data and results. K.R.S. drafted the manuscript. All authors reviewed and edited the manuscript.