Maintenance of Goal Blood Pressure, Cholesterol, and A1C Levels in Veterans With Type 2 Diabetes After Discharge From a Pharmacist-Managed Ambulatory Care Clinic

This was a single-center, retrospective, observational cohort analysis. The primary outcome was length of time to failure of maintenance of a SBP goal of <130 mmHg, an LDL cholesterol goal of <100 mg/dL, or an A1C goal of <7% in patients with type 2 diabetes after being discharged from a pharmacist-managed ambulatory care clinic to receive usual care. The secondary goal of the study was to determine if any factors present at clinic discharge were associated with a shorter time to failure of goal maintenance.

To identify potential risk factors, we collected information on significant demographic and clinical characteristics, including patients’ age, sex, copayment exemption status (i.e., whether a patient was responsible for a copayment on any medication), smoking status, medical comorbidities, and pertinent physical exam findings (i.e., height, weight, BMI, and blood pressure), and laboratory values (i.e., SBP, serum creatinine, LDL cholesterol, total cholesterol, A1C, and microalbumin/creatinine ratio). Comorbid conditions recorded including chronic kidney disease (CKD), CAD, myocardial infarction, coronary artery bypass graft (CABG), percutaneous coronary intervention, cerebrovascular accident/transient ischemic attack, heart failure, chronic obstructive pulmonary disease, peripheral artery disease, and abdominal aortic aneurysm. Information was also recorded regarding patients’ number of prescription medications, insulin use, duration of enrollment in the pharmacy clinic before discharge, and number of visits to the PCP during their clinic enrollment.

The pharmacy-based clinics within the VA St. Louis Health Care System manage patients mostly through a formal electronic consultation from other health care providers within the system. Through this process, most patients are referred to this service for aid in the management of hypertension, hyperlipidemia, diabetes, or a combination thereof. However, patients are also referred for complex medication reconciliation reviews and management of other chronic disease states such as heart failure, chronic obstructive pulmonary disorder, and thyroid disease. Within the VA system, clinic pharmacists have a written scope of practice that includes the ability to order, change, and discontinue noncontrolled medications and to order relevant laboratory tests.

Study patients were identified through a medical informatics database query on the basis of having had an encounter in one of the pharmacy-based clinics between 1 July 2007 and 30 June 2009. During this time frame, patients were managed by pharmacists in six face-to-face clinics and two telephone clinics. Quantifying the precise number of pharmacists involved in these clinics is difficult because some clinic responsibilities were handled by several rotating pharmacist faculty members. Approximately 1,800 patients were seen in the clinics during the study time period and were reviewed for study inclusion.

The records of patients with type 2 diabetes who met initial screening criteria were analyzed to determine whether they met study inclusion criteria. The study included all patients aged 18–89 years who had two consecutive measurements of either SBP, LDL cholesterol, or A1C at goal by discharge and had a documented history of type 2 diabetes, an A1C value ±6.5% during clinic enrollment, or were on any diabetes medication (including insulin, biguanides, secretagogues, thiazolidinediones, GLP-1 receptor agonists, DPP-4 inhibitors, amylin analogs, and α-glucosidase inhibitors).

Patients were excluded if they did not have a follow-up LDL cholesterol or A1C measurement within 1.25 years after discharge or if they had no follow-up SBP measurement within 0.5 years after discharge. For LDL cholesterol, clinical practice guidelines recommend annual lipid panel tests. A follow-up time of 1.25 years was selected to ensure that patients were maintained within goal while allowing a time buffer for patients who might have had to reschedule their clinic appointment, were late in scheduling an annual follow-up appointment, or for whom a provider forgot to order laboratory testing. Although the follow-up criteria could have been shortened for A1C, investigators felt that having a repeat A1C test within 1.25 years of discharge was an adequate time period to ensure reasonable follow-up while again allowing for scheduling or other logistical difficulties. Because SBP is more variable, investigators narrowed the follow-up interval to within 0.5 years to ensure that patients were being followed up in a reasonable amount of time to maintain blood pressure control while balancing the fact that patients might not return to the clinic in a short time span if they were otherwise healthy and having no acute problems.

Patients’ data were collected for a maximum of 3 years after their clinic discharge date. For LDL cholesterol and A1C, time to goal maintenance failure was defined as the time to the first laboratory measurement that was above the given goal. Because of the greater variability in SBP, time to maintenance failure for that goal was defined as the first of three consecutive above-goal readings or documentation of a hypertensive urgency/emergency after clinic discharge.

For data analysis, the primary outcome was analyzed by Kaplan-Meier survival analysis to evaluate the time to maintenance failure for each goal. All covariate data from risk factors were analyzed by Cox proportional hazards to detect significant associations with time to goal maintenance failure. Finally, a step-wise Cox proportional hazard model-building procedure was conducted using forward selection (entering covariates with the smallest P values first and only those statistically significant on individual analysis). P <0.05 was deemed statistically significant.

This study was approved by the institutional review boards of both the VA St. Louis Health Care System and the St. Louis College of Pharmacy.

Sixty-nine patients who were discharged from the clinic after achieving goals but then failed to maintain at least one of the three goals also met the other study criteria. Patient characteristics are summarized in Table 1. Subjects were predominantly elderly, obese men. The mean duration of clinic enrollment was 6.4 months, and patients had approximately two PCP visits (mean 1.8) while enrolled in a clinic. Forty-two patients failed to maintain SBP goals, 35 failed to maintain LDL cholesterol goals, and 22 failed to maintain A1C goals after clinic discharge. Of these patients, 26 failed multiple goals; therefore, these populations are not mutually exclusive and were included in the analyses for each applicable goal.

Subjects’ clinical characteristics are shown in Table 2. With the exception of A1C, patients discharged from a pharmacy-based clinic displayed all-around improvements in their target goals compared to both their baseline values and discharge values for the total study population. Because a clinical pharmacy may not have been consulted to manage all three problems for every patient (i.e., blood pressure, LDL cholesterol, and diabetes), some patients may already have been at goal for some of the three parameters and were being managed only for those not already at goal. For example, some patients with diabetes who were already maintaining their A1C goal were referred to a clinic for cholesterol management only. Table 2 summarizes the overall impact of clinical pharmacy services on SBP, LDL cholesterol, total cholesterol, and A1C for the duration of patients’ clinic enrollment. Clinical characteristics of the patients being managed for each individual goal are shown in Table 3. The percentages of change from baseline to discharge for each clinical characteristic are shown in Tables 2 and 3. The largest changes were observed for total cholesterol (–15%) and LDL cholesterol (–13%) within the total study population. As shown in Table 3, substantial improvements were also seen in the individual groups of those failing to maintain SBP and LDL cholesterol goals (–12% and –27%, respectively).

Forty-two patients who were discharged from a clinic after achieving the SBP goal failed to maintain that goal later. At enrollment, the mean age in this group was 63.8 years old (SD 10.79 years), and the mean SBP was 135.6 mmHg (SD 15.48 mmHg). Figure 1 depicts the time to goal maintenance failure in this group, with a mean time to failure of 9.4 months (SD 8.75 months). As shown in Table 4, factors that were significantly associated with increasing a patient’s rate of failure to maintain the SBP goal included history of CKD (HR 13.16, SD 7.48, P = 0.026), CABG (HR 3.869, SD 3.73, P = 0.008), or heart failure (HR 3.839, SD 3.32, P = 0.019).

Thirty-five patients who were discharged from a clinic after meeting their LDL cholesterol goal later failed to maintain this goal. The mean age at clinic enrollment was 64 years (SD 9.76 years). At discharge, the mean LDL cholesterol was 82 mg/dL (SD 27.45 mg/dL) and mean BMI was 32 kg/m² (SD 6.10 kg/m²), and patients averaged four PCP visits until goal maintenance failure. Depicted in Figure 2, the mean time to failure for the LDL cholesterol goal was 25.8 months (SD 14.5 months). No characteristics were found to be statistically significant risk factors for LDL cholesterol goal failure.

Twenty-two patients who were discharged from a clinic after meeting the A1C goal subsequently failed to maintain that goal. The mean time to failure was 20.4 months (SD 15.1 months). Figure 3 shows the Kaplan-Meier curve associated with this failure. Factors that were significantly associated with a shorter time to failure of the A1C goal are shown in Table 4. Age at clinic enrollment (HR 1.08, SD 0.15, P = 0.015) and A1C at clinic discharge (HR 2.61, SD 2.05, P = 0.028) demonstrated statistical significance. The mean age at clinic enrollment was 63.4 years old (SD 8.39 years) and mean A1C at discharge was 6.42% (SD 0.52%).

Consistent with other studies (2–5, 11–15), the results of this study support the value of clinical pharmacists in disease state management, and specifically in assisting people with type 2 diabetes with improving glycemic control, blood pressure, and cholesterol levels. Within the individual goal failure groups, improvements were seen during clinic enrollment, although the improvement in A1C was modest. Given that the median A1C was 6.5% at clinic enrollment for those in the individual A1C goal failure group, this small improvement to 6.4% at discharge may have been attributable to the fact that many patients had well-controlled glycemia upon enrollment and were referred for hyperlipidemia or hypertension management, in which case there was little room for A1C improvement. On average, patients were taking 13 medications, and ∼40% of patients were using insulin. Polypharmacy and the need for aggressive glycemic therapy may be factors that have some bearing on the difficulty in achieving and sustaining glycemic control in this population.

This study showed that there was generally good durability of maintaining blood pressure, lipid, and glycemic goals in veterans with type 2 diabetes. However, this was less true for SBP, for which goal maintenance durability was ∼9 months, compared to nearly 2 years for LDL cholesterol and A1C goals. Because SBP is likely the most volatile of these parameters, it is not surprising that the related time to goal failure was shorter than those of LDL cholesterol and A1C.

The mean times to goal maintenance failure in this study were comparable to those reported by Pirraglia et al. (17): A1C and LDL cholesterol failures of 21.3 months each and SBP failure of 6.9 months. SBP goal maintenance was slightly more durable in the present study than in the study by Pirraglia et al., given that our definition of SBP goal failure was stricter, requiring three consecutive above-goal readings versus a single above-goal reading.

Although this study did not find a significant association between baseline SBP and time to maintenance failure for that goal, patients with a history of CKD, heart failure, or CABG had a significantly shorter time to SBP goal failure. Because the definition for failure of this goal differed from that in the study by Pirraglia et al., it is not surprising that the factors significantly affecting time to goal maintenance failure also varied.

Given the marginal stability of SBP, the findings of this study indicate that hypertensive veterans with type 2 diabetes, and particularly those with a history of CKD, heart failure, or CABG, may benefit from continued follow-up with a pharmacist-managed clinic after achieving their goal instead of being discharged to usual care. Based on the study by Pirraglia et al. (17), the same consideration could also be given to patients with an elevated SBP level at the time of clinic enrollment. However, that study did not identify a specific SBP level that would be considered elevated and a risk for a shorter time to goal failure.

Although Pirraglia et al. identified insulin use as a risk factor for shorter time to LDL cholesterol goal maintenance failure, the present study found no similar risk factors. However, it did find that patients of advanced age or having higher A1C levels at discharge may be at higher risk for failure to maintain their A1C goal after clinic discharge. Because the study by Pirraglia et al. identified only BMI as a significant risk factor for time to A1C goal maintenance failure (17), the results of these two studies differ in terms of risk factors for these two outcomes. Although these factors may contribute to a shorter time to failure of maintaining an LDL cholesterol or A1C goal, most patients, including those with these potential risk factors, were able to maintain both goals for ∼2 years. This should be kept in mind when considering whether to forgo clinic discharge in favor of continuing clinic follow-up for patients with these characteristics.

This study had some limitations. It included only patients with diabetes who had failed to maintain their clinical goals within 3 years of discharge from the pharmacy-based clinics. There was also a relatively small sample size; only 69 patients met study criteria to be included in the statistical analyses. This was ∼30% of the population size in the study by Pirraglia et al., which may have contributed to the lack of accord in the risk factors identified in the two studies (17). In addition, it may have been helpful for this study to have a comparison group of patients with type 2 diabetes who maintained their clinical goals for 3 years after being discharged from the pharmacy clinic. Also, because this study was retrospective, it was not possible to control for all confounding factors that may have influenced its findings. Factors such as visits to other specialty providers, medication adherence, appointment adherence, diet, physical activity level, and patient literacy were not documented and could have potentially influenced the results. Interpretation of the results for SBP in this study is complicated by the fact that, in 2013, ADA changed its recommended SBP goal for people with diabetes from <130 mmHg to <140 mmHg (18). Finally, because this study was conducted on veterans with type 2 diabetes receiving care through a VA health care system, its findings may not be fully applicable to other patient populations.

In conclusion, this study provides further evidence that pharmacists can have a significant impact as part of the health care team in improving patient care in the diabetes patient population, and specifically in helping these patients achieving blood pressure, cholesterol, and glycemic goals. From these results, it can be concluded that, once achieved in a pharmacy-based clinic, LDL cholesterol and A1C goals are relatively durable, although the SBP goal does not have the same durability.

This study also adds new evidence that many patients may be safely discharged from a disease state management clinic and that resources can then be utilized to target those with specific risk factors for failure to maintain their clinical goals. These results may alter current practices and criteria for patient discharge, especially for patients who achieve their SBP goals but may still benefit from additional follow-up in the pharmacy-based clinic (e.g., patients with type 2 diabetes and CKD). Because the mean time to goal maintenance failure for SBP was 9.4 months, scheduling a follow-up appointment for 6–9 months after discharge seems reasonable. Additional research into the factors that encourage maintenance of blood pressure, cholesterol, and glycemic goals and reduce cardiovascular risk will be vital in determining optimal strategies for management of these chronic conditions.

No potential conflicts of interest relevant to this article were reported.