OBJECTIVE—To compare health resource utilization in patients with diabetes between the U.S. and Canada.
RESEARCH DESIGN AND METHODS—We combined measures of health care utilization, personal, demographic, health status, functional status, and comorbid conditions from the current National Health Interview Survey (U.S.) and the National Public Health Survey (Canada). A binary logistic regression analysis was used to examine how country of residence influences the probability of accessing health care resources in adult Caucasian subjects after controlling for potential confounders.
RESULTS—Subjects from Canada (n = 521) were more likely to have contact with a general physician (odds ratio [OR] 4.01, 95% CI 2.26–7.14), eye specialists (1.46, 1.08–1. 98), and any physician (3.02, 1.03–8.84) in the past year than their American counterparts (n = 825) but were less likely to have had contact with other medical specialists (0.33, 0.24–0.46). Subjects in Canada were also more likely to have been hospitalized overnight (1.79, 1.17–2.75) or to have contact with a health care professional in the previous 12 months (3.35, 1.01–12.81).
CONCLUSIONS—Significant disparities exist in health service utilization for adult Caucasian individuals with diabetes in Canada versus the U.S. after controlling for various confounders. From what is known regarding optimal treatment of diabetes, those with diabetes in the U.S. have a greater chance of not receiving recommended care.
There has been a great deal of interest in recent years regarding differences in health service utilization patterns in Canada and the U.S. Since the emergence of a population-wide health insurance system in Canada in the early 1960s, the two systems that were once quite similar have evolved in very different directions. The interest in comparisons comes from both sides of the border. In the U.S., there is continuing discussion regarding population-wide health insurance coverage, as exists in Canada. In Canada, there is a growing interest in health reforms surrounding managed care, which has evolved in the U.S. A central consideration is how differences in health care system organization and delivery influence utilization, with its potential impact on outcomes and costs. Those with diabetes form an excellent set of subjects for examining this issue, as this chronic medical disorder is associated with considerable morbidity and mortality that is known to be influenced by intensive medical intervention (1–3).
Previous health resource utilization comparisons at the population level indicate that Canadians have higher hospital utilization rates by 10–30% (4,5), and when stratified by income for both the U.S. and Canada, Canadians have more physician visits per person annually for all except those in the highest income quintile (6). Specific subpopulations have shown varied results depending on the group of interest and specific utilization measure of interest. For example, rates of physician use for the elderly are higher in Canada than in the U.S (7), while use of invasive procedures and interventions in the treatment of cardiovascular disease are consistently higher in the U.S (8–12). Only one study has examined utilization in those with diabetes, in which a higher frequency of physician visits in Canada compared with the U.S. was found for a relatively young healthy population with type 1 diabetes (13).
In this analysis, we compare resource utilization between the U.S. and Canada in Caucasian adult subjects with diabetes by examining data obtained from two national surveys. The primary aim of this study was to determine whether differences in health care resource utilization exist between Canadian and American individuals with diabetes after controlling for confounding factors.
RESEARCH DESIGN AND METHODS
Subjects for this analysis were obtained from the 1999 National Health Interview Survey (NHIS) (14,15), conducted by the National Center for Health Statistics and the Centers for Disease Control and Prevention, and the 1998 National Population Health Survey (NPHS) household component, conducted by Statistics Canada (16). These surveys included ∼30,800 and 17,000 respondents, respectively, drawn from community-dwelling individuals and households in both countries. Both use a complex sample design and collect data regarding personal and demographic characteristics, self-reports of medical disorders diagnosed by a health care professional, functional status, and the use of health resources.
A series of binary logistic regression equations based on the demand model (17) were created, with a measure of health care resource utilization as the dependent variable. The dependent binary variable indicates any contact in the previous year with a health professional or contacts with a medical generalist, specialist, or eye specialist, as well as composite physician contact and overnight hospitalization. Explanatory variables known to influence health resource consumption include age (by 5-year intervals, from age 20–24 to 80+ years), sex (male/female), education level (no postsecondary or at least some postsecondary), location (urban/rural) (18), household income in either U.S. or Canadian currency (≥$20,000/<$20,0,000), and family structure (spouse or partner, two adults with child(ren), single parent with child(ren), single living alone, or single living with other adults) (18–23). Health indicator variables included general health status (excellent, very good, good/fair, or poor), likelihood of major depression >80% based on the Composite International Diagnostic Interview (24–26) (yes/no), independence in activities of daily living (ADLs) (yes/no), treatment for diabetes (no pharmacological treatment, oral hypoglycemic agents only, insulin only, or both oral hypoglycemic agents and insulin), presence of heart disease (yes/no), and presence of hypertension (yes/no) (27–29). No attempt was made to differentiate those variables with respect to appropriate or inappropriate drivers of health resource consumption. Additionally, a dummy variable indicating Canada or U.S. residence was included to capture differences in delivery of service by country.
To facilitate the merging of data from the NHIS and the NPHS, variables were recoded such that similar information could be captured. The survey questions used to capture all dependent variables and explanatory variables where survey wording differences exist are shown in Table 1. To facilitate meaningful comparisons, only a well-defined subgroup from each survey was included. Adults aged ≥20 years, those with a self-report of a diagnosis of diabetes from a health care professional, and those of white race defined the study population. Only those respondents with complete information regarding the variables of interest were included. Correlations between explanatory variables were obtained to assess for model multicollinearity, and testing for interaction between country of origin and income (5,6) was assessed for each model.
All statistical tests were performed using SPSS for Windows version 11.0.1 (31) or SUDAAN version 8.0 (32). Although each survey requires unique handling to account for multistage sampling and nonequal probabilities of selection for the survey to determine population estimates, these were not incorporated in the regression analysis. The rationale for handling the data in this manner is discussed further in the text. Statistical significance was defined by a P value <0.05.
RESULTS
The characteristics of white adult individuals with diabetes who provided complete information for all explanatory variables are compared by country of origin in Table 2. A total of 1,346 subjects were included, with 521 from Canada and 825 from the U.S. Statistically significant differences using the χ2 test exist between the two sample populations for all variables except for sex, education level, major depression, and health professional visits. These differences were attenuated when analyzed by age group. The discrepancy in age distribution persisted even when population estimates were obtained with the use of appropriate weighting and stratification. The two sample populations were similar with respect to those treated and not treated with pharmacological therapy (χ2 = 0.92) for their diabetes. Many explanatory variables were correlated with one another. No interaction between income and country was detected.
The results for the six equations are given in Table 3. In general, the presence of heart disease, hypertension, and the use of pharmacological therapy to treat diabetes have odds ratios (ORs) of utilization >1 for all equations. Overnight hospitalization is associated with requiring assistance with ADLs and a fair or poor general health status.
Those subjects who reside in Canada are much more likely to have contact with a general physician or eye specialist in the previous 12 months than their American counterparts. In contrast, Canadians are much less likely to have contact with a specialist physician (other than eye specialists) in the preceding year. The OR of aggregate physician contact in the previous 12 months is 3.02 (P = 0.044) for those subjects in Canada compared with those in the U.S.
Subjective analysis of the disparity in survey questions between the NPHS and the NHIS suggests that there may be a slight bias toward higher reported contact with physicians in the U.S., as the Canadian survey excludes contacts as overnight patients. Furthermore, due to question differences, it was impossible to completely separate specialist physician from allied health professional for mental health care. However, the base analysis for specialist physician contact was unchanged with the alternate incorporation of this information, which had a similar OR of use by country of 0.31 (95% CI 0.22–0.44).
Those individuals with diabetes who reside in Canada have an OR of health professional consultation of 3.35. Subjective analysis of the survey questions for health care professional utilization suggests that there may be slight bias toward higher use in the U.S., as the Canadian question excludes contacts as overnight patients. A bias toward higher utilization in the U.S. may also be present, as utilization includes services by a respiratory therapist and midwife, which are not specifically included in the NPHS. Both of these differences are likely to have only minor effects.
A significant association for country of residence was found with respect to hospitalization, with those from Canada having an overnight hospitalization in the past 12 months. Subjective interpretation of the survey questions suggest that hospital utilization may be biased upwards in the Canada survey, as it does not exclude emergency room–only stays and specifically mentions nursing home and convalescent home stays as well as hospital stays.
In order to explore the effect of being uninsured in the U.S. in this model, those who answered “no” to the NHIS question “Are you covered by health insurance or some other kind of health care plan?” were identified. In this sample, 23.2% were without insurance, which is somewhat higher than the 13.5% previously reported (33). A reanalysis excluding those without insurance yielded no significant change in results, with the OR of resource utilization by country remaining of similar magnitude, direction, and statistical significance.
CONCLUSIONS
We combined observations from Canadian and U.S. population health surveys for individuals with diabetes and developed an analysis to examine differences in utilization between the two nations. The results of this analysis suggest that marked differences in delivery of care exist for Caucasian adults with diabetes between the U.S. and Canada. Specific focus on physician use reveals a higher probability of visiting any physician in the past year for those in Canada versus the U.S. Canadian individuals with diabetes are more likely to have seen an eye specialist and to have seen a general practitioner in the past year, while those in the U.S. are more likely to have seen a medical specialist. Any overnight hospital stay or contact with a health professional in the previous year was similarly greater for those subjects in Canada compared with the U.S.
Certain caveats merit consideration in the interpretation of these results. Although this analysis uses national surveys designed to represent entire populations, this analysis does not compare health resource utilization from a population standpoint. Our rationale for using the data in this way was to create a broad sample from both the U.S. and Canada with similar characteristics in order to approximate an “iso-need” group from each country, which facilitates the interpretation of comparative resource use. Furthermore, the use of weighting to represent population size may not always be appropriate (34). After selecting only white adult subjects with diabetes, differences remained in the Canadian and U.S. samples, which were controlled for by inclusion into the regression equation. However, similar to any analysis using this technique, we cannot be certain that all differences are adequately controlled for.
Further consideration must be made to the process of combining data from two distinct surveys. Many of the survey questions of interest in the NPHS and NHIS were very similar and likely captured similar information. However, subtle wording differences and the disparity in the number of distinct questions in each survey used to capture the information may lead to systematic bias, which was subjectively assessed for each dependent variable. For hospital utilization, the bias is in the direction of increased use in Canada, as it captures overnight stays in the emergency room and nursing and convalescent homes, although this may be minor, as the surveys were conducted on those dwelling in the community. However, we cannot be certain that the increased OR is fully or partially a result of differences in wording. Health care professional and physician utilization questions are quite similar, although there may be a slight bias toward lower utilization in Canada, as it excludes contacts that occur within the hospital, whereas the NHIS does not. This may strengthen the finding of increased utilization, which was found for all but the medical specialist group in Canada.
Caution should be used when interpreting the variance of the OR of the explanatory variables. Multicollinearity was detected in our model, which leads to larger variances and wider CIs. We felt that it was important to include all of the potential variables influencing health resource utilization in order to increase the confidence that differences between countries were not a result of confounding, while recognizing that many accepted drivers of utilization such as depression (29) may not be significant in our model due to multicollinearity.
Considerable differences in age distribution for those with diabetes exist in our sample. To determine whether this was a result of sampling differences between the two surveys, the age distribution for all of those with diabetes was computed using weighting and stratification information from each survey to obtain population estimates. These differences persist in a similar pattern to the distribution in Table 2. We are aware of no previously reported differences in the age distribution of those with diabetes in Canada versus the U.S. For our analysis, we attempted to control for this disparity by using the smallest age interval possible.
Despite the previous cautions in interpreting these findings, our results are very consistent with many previous analyses of health resource utilization in Canada and the U.S. Overall higher rates in physician use adjusted for age, sex, and morbidity have been found for all but the highest income quintiles in Canada versus the U.S (6), similar to other reports of general population utilization of physician resources (4,35). Selected populations have revealed similar trends in physician visits, including the elderly (7). Canadians post-myocardial infarction have been found to have a greater number of generalist visits but fewer specialist visits than their American counterparts (36). Hospital utilization has been reported to be greater in Canada than in the U.S. for both the general population (4,5) and for subgroups, such as those with systemic lupus erythematosus (37). To our knowledge, the only study analyzing resource utilization in those with diabetes is by Booth et al. (13), who documented a higher frequency of outpatient visits in Canadian versus the U.S. in those subjects enrolled in the Diabetes Control and Complications Trial (DCCT). The population included those <40 years of age with type 1 diabetes only who were free of any significant complications. Our results generalize these findings to Caucasian adults with diabetes, including those with complications, and add further detail in terms of specific physician type, other health care professional use, and hospital stays in a much less selected population.
Speculation can be given as to the drivers of the disparity in the health resource utilization found here. The gate-keeping system where access to specialist care is controlled by a generalist in Canada is likely to influence the disparity in generalist and specialist use between these two nations, along with the differences in proportions of generalists to specialists and differences in “tastes” for care in these two nations. Moral hazard is likely playing a significant role, as a greater proportion of those in the U.S. are uninsured and face the full costs of medical care or pay a copayment compared with those in Canada, where universal insurance is provided. This may partially explain higher physician and health care contacts in Canada. However, we did not find that the uninsured individuals in the U.S. were the sole driving force for this difference, as the results were unchanged when only those with some type of insurance were analyzed. This would suggest that future policy focusing only on uninsured individuals may be too narrowly focused. Unfortunately, specifics of insurance coverage including comprehensiveness and copayment information was not available to further explore insurance coverage and resource utilization.
This study raises important questions with regards to how care is delivered to patients with diabetes. Our outcome used was dichotomous, indicating absence or presence of resource utilization in the past year, not quantity of resources utilized. As such, it provides a starker comparison considering that results from the DCCT (1) and U.K. Prospective Diabetes Study (UKPDS) (2,3) suggest that tighter control through more intensive and frequent health care contacts is known to be associated with improved outcomes. This improved glycemic control attained with intensive therapy has been demonstrated to result in significant cost savings within 2 years despite the initial increased cost of management (38). While our data does not capture information regarding control of diabetes, it is a reasonable assumption that not meeting with a health professional for a year represents suboptimal care.
Despite the fact that per capita health expenditure in the U.S. during the time period of observations was $4,178 (U.S.$) compared with $2,285 (U.S.$) in Canada (39), we found that those with diabetes in the U.S. are more likely to not receive recommended care. Furthermore, our findings regarding physician care indicate that individuals with diabetes receive very different types of care in the two countries, one with a greater reliance on general medicine (Canada) and one with a greater emphasis on specialized medicine (U.S.). Reports from the literature raise questions about the efficacy of either approach but do not provide definitive answers (40).
In summary, we have documented disparity in delivery of health care to Caucasian adults with diabetes between the U.S. and Canada. After controlling for potential confounders, those with diabetes are more likely to be hospitalized and are more likely to have seen a general practitioner or eye specialist than those in the U.S., while they are less likely to have used the services of a medical specialist. Due to the nature of this analysis, these results are hypothesis generating but are consistent with other reports in the literature. Further investigation of the contributing factors leading the disparity of resource utilization and the effect on clinical outcomes and health care costs should be undertaken.
Survey questions from the NHIS and NPHS used to derive dependent variables of health resource consumption and selected dependent variables
| Variable . | NHIS (U.S.) . | NPHS (Canada) . |
|---|---|---|
| Core question | During the past 12 months, have you seen or talked to any of the following health care providers about your own health? | Not counting when you were an overnight patient, in the past 12 months, how many times have you seen or talked on the telephone with the following about your physical, emotional or mental health? |
| General practitioner | …a general doctor who treats a variety of illness (general practice, family medicine, internal medicine)? | …family doctor or general practitioner? |
| Medical specialist | …a medical doctor who specialized in a particular medical disease or problem (other than obstetrician gynaecologist, psychiatrist, or ophthalmologist? …a doctor who specializes in women’s health (an obstetrician/gynaecologist)? | …other medical doctor (such as a surgeon, allergist, orthopaedist, gynaecologist or psychiatrist)? |
| Eye doctor | …an optometrist, optician or eye doctor? | …eye specialist (such as an ophthalmologist or optometrist) |
| Health care professional | …mental health professional such as a psychiatrist, psychologist, psychiatric nurse or clinical social worker; optometrist, optician or eye doctor; chiropractor; physical therapist, speech therapist, respiratory therapist, audiologist, or occupation therapist; nurse practitioner, physician assistant or midwife; doctor who specialized in women’s health (obstetrician/gynaecologist); a medical doctor who specializes in a particular medical disease or problem; a general doctor who treats a variety of illnesses? | …family doctor or general practitioner; eye specialist (such as an ophthalmologist or optometrist); other medical doctor (such as surgeon, allergist, orthopaedist, gynaecologist or psychiatrist); chiropractor, physiotherapist, social worker or counsellor, psychologist; speech, audiolegist or occupational therapist. |
| Overnight hospital | During the past 12 months were you a patient in a hospital overnight (not including an overnight stay in the emergency room)? | In the past 12 months, have you been a patient overnight in a hospital, nursing home or convalescent home? |
| Medical conditions | Have you ever been told by a doctor or health professional that you have | We are interested in “long-term conditions” that have lasted or are expected to last 6 months or more that have been diagnosed by a health professional. Do you have: |
| Diabetes | …diabetes or sugar diabetes? | …diabetes? |
| Heart disease | …coronary artery disease? angina or angina pectoris? a heart attack? any kind of heart condition or heart disease? congestive heart failure? | …heart disease? |
| Hypertension | Were you told on two or more different visits that you had hypertension, also called high blood pressure? | …high blood pressure? |
| Difficulties with activities of daily living? | Because of a physical, mental, or emotional problem, do you need the help of other persons with personal care needs, such as eating, bathing, dressing, or getting around inside the home? | Because of any condition or health problem, do you need the help of another person in: personal care such as washing, dressing or eating; in moving about inside the house? |
| Treatment modality for diabetes | Are you now taking insulin? Are you now taking diabetic pills to lower your blood sugar? These are sometimes called oral agents or oral hypoglycaemic agents. | Do you take insulin for this? Do you take any other treatment or medication for this? What kind of treatment or medication? (Drug/Diet/Exercise or physiotherapy/other) |
| Rural or urban residence | Reside in Metropolitan Statistical Area (MSA) defined as one city of ≥50,000 or urbanized area of ≥50000 and total metropolitan population of at least 100000 or in surrounding community with relevant commuting patterns to a central city of ≥250000, as well as other criteria (18) | Urban: Defined based on residing in a continuously built-up areas having a population concentration of 1,000 or more and a population density of 400 or more per square kilometre based on the previous census; or very large urban area, together with adjacent urban and rural areas which have a high degree of economic and social integration with that urban area, with a core population of ≥100000 |
| Variable . | NHIS (U.S.) . | NPHS (Canada) . |
|---|---|---|
| Core question | During the past 12 months, have you seen or talked to any of the following health care providers about your own health? | Not counting when you were an overnight patient, in the past 12 months, how many times have you seen or talked on the telephone with the following about your physical, emotional or mental health? |
| General practitioner | …a general doctor who treats a variety of illness (general practice, family medicine, internal medicine)? | …family doctor or general practitioner? |
| Medical specialist | …a medical doctor who specialized in a particular medical disease or problem (other than obstetrician gynaecologist, psychiatrist, or ophthalmologist? …a doctor who specializes in women’s health (an obstetrician/gynaecologist)? | …other medical doctor (such as a surgeon, allergist, orthopaedist, gynaecologist or psychiatrist)? |
| Eye doctor | …an optometrist, optician or eye doctor? | …eye specialist (such as an ophthalmologist or optometrist) |
| Health care professional | …mental health professional such as a psychiatrist, psychologist, psychiatric nurse or clinical social worker; optometrist, optician or eye doctor; chiropractor; physical therapist, speech therapist, respiratory therapist, audiologist, or occupation therapist; nurse practitioner, physician assistant or midwife; doctor who specialized in women’s health (obstetrician/gynaecologist); a medical doctor who specializes in a particular medical disease or problem; a general doctor who treats a variety of illnesses? | …family doctor or general practitioner; eye specialist (such as an ophthalmologist or optometrist); other medical doctor (such as surgeon, allergist, orthopaedist, gynaecologist or psychiatrist); chiropractor, physiotherapist, social worker or counsellor, psychologist; speech, audiolegist or occupational therapist. |
| Overnight hospital | During the past 12 months were you a patient in a hospital overnight (not including an overnight stay in the emergency room)? | In the past 12 months, have you been a patient overnight in a hospital, nursing home or convalescent home? |
| Medical conditions | Have you ever been told by a doctor or health professional that you have | We are interested in “long-term conditions” that have lasted or are expected to last 6 months or more that have been diagnosed by a health professional. Do you have: |
| Diabetes | …diabetes or sugar diabetes? | …diabetes? |
| Heart disease | …coronary artery disease? angina or angina pectoris? a heart attack? any kind of heart condition or heart disease? congestive heart failure? | …heart disease? |
| Hypertension | Were you told on two or more different visits that you had hypertension, also called high blood pressure? | …high blood pressure? |
| Difficulties with activities of daily living? | Because of a physical, mental, or emotional problem, do you need the help of other persons with personal care needs, such as eating, bathing, dressing, or getting around inside the home? | Because of any condition or health problem, do you need the help of another person in: personal care such as washing, dressing or eating; in moving about inside the house? |
| Treatment modality for diabetes | Are you now taking insulin? Are you now taking diabetic pills to lower your blood sugar? These are sometimes called oral agents or oral hypoglycaemic agents. | Do you take insulin for this? Do you take any other treatment or medication for this? What kind of treatment or medication? (Drug/Diet/Exercise or physiotherapy/other) |
| Rural or urban residence | Reside in Metropolitan Statistical Area (MSA) defined as one city of ≥50,000 or urbanized area of ≥50000 and total metropolitan population of at least 100000 or in surrounding community with relevant commuting patterns to a central city of ≥250000, as well as other criteria (18) | Urban: Defined based on residing in a continuously built-up areas having a population concentration of 1,000 or more and a population density of 400 or more per square kilometre based on the previous census; or very large urban area, together with adjacent urban and rural areas which have a high degree of economic and social integration with that urban area, with a core population of ≥100000 |
Characteristics of the sample population by country
| Characteristic . | Canada (%) . | US (%) . | χ2 . |
|---|---|---|---|
| n | 521 | 825 | 0.000* |
| Age in years | |||
| 20–24 | 0.4 | 11.0 | |
| 25–29 | 0.8 | 10.7 | |
| 30–34 | 2.5 | 10.2 | |
| 35–39 | 3.6 | 13.0 | |
| 40–44 | 4.8 | 10.5 | |
| 45–49 | 6.7 | 10.2 | |
| 50–54 | 6.0 | 7.6 | |
| 55–59 | 10.4 | 5.8 | |
| 60–64 | 10.6 | 5.6 | |
| 65–69 | 13.6 | 4.0 | |
| 70–74 | 15.2 | 3.3 | |
| 75–79 | 13.4 | 4.0 | |
| 80+ | 12.1 | 4.1 | |
| Female sex | 51.6 | 54.3 | 0.339 |
| No postsecondary education | 58.9 | 56.4 | 0.355 |
| Family structure | 0.000* | ||
| Spouse or partner | 37.0 | 23.9 | |
| Two adults with child(ren) | 15.4 | 42.5 | |
| Single parent with child(ren) | 4.4 | 4.6 | |
| Single living alone | 37.4 | 10.4 | |
| Single living with other adults | 5.8 | 18.5 | |
| Household income ≥$20,000 | 59.1 | 73.7 | 0.000* |
| Urban residence | 41.7 | 83.0 | 0.000* |
| Good to excellent general health state | 60.8 | 88.6 | 0.000* |
| Independent with ADL | 88.1 | 98.2 | 0.000* |
| Depression | 6.7 | 8.0 | 0.384 |
| Hypertension | 44.5 | 57.6 | 0.000* |
| Heart disease | 21.9 | 33.2 | 0.000* |
| Treatment of diabetes | 0.000* | ||
| No pharmacologic therapy | 26.3 | 26.5 | |
| Oral hypoglycemics only | 49.9 | 41.1 | |
| Insulin only | 18.8 | 20.8 | |
| Both oral hypoglycemics and insulin | 5.0 | 11.5 | |
| All physician services | 98.1 | 93.5 | 0.000* |
| General practitioner use | 95.4 | 85.3 | 0.000* |
| Medical specialist use | 42.4 | 54.1 | 0.000* |
| Eye doctor use | 62.6 | 54.8 | 0.002* |
| Overnight hospital stay | 21.5 | 9.2 | 0.000* |
| Health professional visit | 98.5 | 97.2 | 0.301 |
| Characteristic . | Canada (%) . | US (%) . | χ2 . |
|---|---|---|---|
| n | 521 | 825 | 0.000* |
| Age in years | |||
| 20–24 | 0.4 | 11.0 | |
| 25–29 | 0.8 | 10.7 | |
| 30–34 | 2.5 | 10.2 | |
| 35–39 | 3.6 | 13.0 | |
| 40–44 | 4.8 | 10.5 | |
| 45–49 | 6.7 | 10.2 | |
| 50–54 | 6.0 | 7.6 | |
| 55–59 | 10.4 | 5.8 | |
| 60–64 | 10.6 | 5.6 | |
| 65–69 | 13.6 | 4.0 | |
| 70–74 | 15.2 | 3.3 | |
| 75–79 | 13.4 | 4.0 | |
| 80+ | 12.1 | 4.1 | |
| Female sex | 51.6 | 54.3 | 0.339 |
| No postsecondary education | 58.9 | 56.4 | 0.355 |
| Family structure | 0.000* | ||
| Spouse or partner | 37.0 | 23.9 | |
| Two adults with child(ren) | 15.4 | 42.5 | |
| Single parent with child(ren) | 4.4 | 4.6 | |
| Single living alone | 37.4 | 10.4 | |
| Single living with other adults | 5.8 | 18.5 | |
| Household income ≥$20,000 | 59.1 | 73.7 | 0.000* |
| Urban residence | 41.7 | 83.0 | 0.000* |
| Good to excellent general health state | 60.8 | 88.6 | 0.000* |
| Independent with ADL | 88.1 | 98.2 | 0.000* |
| Depression | 6.7 | 8.0 | 0.384 |
| Hypertension | 44.5 | 57.6 | 0.000* |
| Heart disease | 21.9 | 33.2 | 0.000* |
| Treatment of diabetes | 0.000* | ||
| No pharmacologic therapy | 26.3 | 26.5 | |
| Oral hypoglycemics only | 49.9 | 41.1 | |
| Insulin only | 18.8 | 20.8 | |
| Both oral hypoglycemics and insulin | 5.0 | 11.5 | |
| All physician services | 98.1 | 93.5 | 0.000* |
| General practitioner use | 95.4 | 85.3 | 0.000* |
| Medical specialist use | 42.4 | 54.1 | 0.000* |
| Eye doctor use | 62.6 | 54.8 | 0.002* |
| Overnight hospital stay | 21.5 | 9.2 | 0.000* |
| Health professional visit | 98.5 | 97.2 | 0.301 |
P < 0.05.
Physician, health care professional and hospital utilization in past 12 months
| Characteristic . | Generalist . | Medical specialist* . | Eye specialist . | All physicians . | Health professional visits . | Overnight hospital stay . |
|---|---|---|---|---|---|---|
| n | 1,337 | 1,153 | 1,337 | 1,311 | 1,340 | 1,346 |
| Age in years (by 5-year increments) | 1.01 (0.94–1.07) | 1.00 (0.95–1.05) | 1.01 (0.97–1.05) | 1.00 (0.87–1.15) | 1.04 (0.89–1.21) | 1.06 (0.99–1.13) |
| Male sex (ref: female) | 1.00 (0.69–1.45) | 0.98 (0.76–1.27) | 0.88 (0.68–1.09) | 0.56 (0.26–1.22) | 0.43 (0.18–1.01) | 0.81 (0.57–1.14) |
| No postsecondary education (ref: some postsecondary) | 0.88 (0.59–1.30) | 0.68 (0.52–0.89) | 0.86 (0.68–1.09) | 0.71 (0.31–1.63) | 0.70 (0.29–1.71) | 0.96 (0.68–1.37) |
| Family structure (ref: spouse or partner) | ||||||
| Two adults with child(ren) | 0.88 (0.52–1.51) | 0.93 (0.64–1.34) | 0.88 (0.64–1.22) | 0.93 (0.31–2.80) | 0.82 (0.24–2.79) | 1.40 (0.84–2.33) |
| Single parent with child(ren) | 0.83 (0.30–2.26) | 1.70 (0.85–3.40) | 0.86 (0.48–1.56) | 0.45 (0.08–2.52) | 0.28 (0.05–1.76) | 1.66 (0.75–3.68) |
| Single living alone | 0.53 (0.29–0.95) | 0.86 (0.59–1.24) | 1.10 (0.78–1.56) | 0.36 (0.12–1.09) | 0.23 (0.07–0.82) | 0.90 (0.55–1.48) |
| Single living with other adults | 0.69 (0.51–1.79) | 0.95 (0.61–1.49) | 0.84 (0.58–1.24) | 3.75 (0.44–32.29) | 3.08 (0.34–28.22) | 1.39 (0.79–2.47) |
| Household income (ref: ≥$20,000) | 1.27 (0.79–2.02) | 0.90 (0.66–1.21) | 0.82 (0.63–1.08) | 0.98 (0.39–2.47) | 1.14 (0.41–3.15) | 1.34 (0.90–1.98) |
| Urban location of residence (ref: rural) | 0.83 (0.51–1.33) | 1.03 (0.77–1.37) | 1.01 (0.77–1.32) | 1.03 (0.41–2.55) | 0.91 (0.33–2.47) | 0.87 (0.60–1.26) |
| Fair to poor general health state (ref: good to excellent) | 0.94 (0.54–1.65) | 1.23 (0.88–1.71) | 1.01 (0.77–1.31) | 1.48 (0.41–5.37) | 1.08 (0.29–4.02) | 1.88 (1.28–2.77) |
| Independent with ADL (no) (ref: yes) | 1.46 (0.42–5.09) | 0.83 (0.48–1.41) | 0.52 (0.31–0.87) | 258.82 (†) | 184.87 (†) | 2.65 (1.54–4.55) |
| Depression (yes) (ref: none) | 1.29 (0.68–2.44) | 0.77 (0.48–1.23) | 0.70 (0.45–1.08) | 1.30 (0.36–4.75) | 0.92 (0.20–4.26) | 1.22 (0.64–2.33) |
| Heart disease present (ref: absent) | 1.39 (0.90–2.13) | 1.78 (1.32–2.39) | 1.38 (1.06–1.78) | 5.53 (1.25–23.61) | 5.25 (1.17–23.62) | 1.56 (1.09–2.22) |
| Hypertension present (ref: absent) | 1.45 (0.99–2.11) | 1.17 (0.90–1.52) | 1.20 (0.95–1.51) | 1.73 (0.76–3.90) | 1.59 (0.67–3.78) | — |
| Treatment of diabetes (ref: no pharmacologic therapy) | ||||||
| Oral hypoglycemics only | 1.77 (1.14–2.76) | 1.34 (0.98–1.83) | 1.38 (1.05–1.81) | 3.67 (1.52–8.81) | 4.81 (1.81–12.81) | 1.37 (0.89–2.11) |
| Insulin only | 1.12 (0.69–1.84) | 1.53 (1.05–2.22) | 1.86 (1.33–2.60) | 4.47 (1.26–15.90) | 7.26 (1.60–33.00) | 1.86 (1.14–3.04) |
| Both oral hypoglycemics and insulin | 1.88 (0.90–3.92) | 1.12 (0.68–1.83) | 2.20 (1.41–3.45) | 6.53 (0.83–51.67) | 6.80 (0.85–54.21) | 0.73 (0.33–1.62) |
| Country (Canada) (ref: U.S.) | 4.01 (2.26–7.14) | 0.33 (0.24–0.46) | 1.46 (1.08–1.98) | 3.02 (1.03–8.84) | 3.35 (1.01–12.81) | 1.79 (1.17–2.75) |
| Characteristic . | Generalist . | Medical specialist* . | Eye specialist . | All physicians . | Health professional visits . | Overnight hospital stay . |
|---|---|---|---|---|---|---|
| n | 1,337 | 1,153 | 1,337 | 1,311 | 1,340 | 1,346 |
| Age in years (by 5-year increments) | 1.01 (0.94–1.07) | 1.00 (0.95–1.05) | 1.01 (0.97–1.05) | 1.00 (0.87–1.15) | 1.04 (0.89–1.21) | 1.06 (0.99–1.13) |
| Male sex (ref: female) | 1.00 (0.69–1.45) | 0.98 (0.76–1.27) | 0.88 (0.68–1.09) | 0.56 (0.26–1.22) | 0.43 (0.18–1.01) | 0.81 (0.57–1.14) |
| No postsecondary education (ref: some postsecondary) | 0.88 (0.59–1.30) | 0.68 (0.52–0.89) | 0.86 (0.68–1.09) | 0.71 (0.31–1.63) | 0.70 (0.29–1.71) | 0.96 (0.68–1.37) |
| Family structure (ref: spouse or partner) | ||||||
| Two adults with child(ren) | 0.88 (0.52–1.51) | 0.93 (0.64–1.34) | 0.88 (0.64–1.22) | 0.93 (0.31–2.80) | 0.82 (0.24–2.79) | 1.40 (0.84–2.33) |
| Single parent with child(ren) | 0.83 (0.30–2.26) | 1.70 (0.85–3.40) | 0.86 (0.48–1.56) | 0.45 (0.08–2.52) | 0.28 (0.05–1.76) | 1.66 (0.75–3.68) |
| Single living alone | 0.53 (0.29–0.95) | 0.86 (0.59–1.24) | 1.10 (0.78–1.56) | 0.36 (0.12–1.09) | 0.23 (0.07–0.82) | 0.90 (0.55–1.48) |
| Single living with other adults | 0.69 (0.51–1.79) | 0.95 (0.61–1.49) | 0.84 (0.58–1.24) | 3.75 (0.44–32.29) | 3.08 (0.34–28.22) | 1.39 (0.79–2.47) |
| Household income (ref: ≥$20,000) | 1.27 (0.79–2.02) | 0.90 (0.66–1.21) | 0.82 (0.63–1.08) | 0.98 (0.39–2.47) | 1.14 (0.41–3.15) | 1.34 (0.90–1.98) |
| Urban location of residence (ref: rural) | 0.83 (0.51–1.33) | 1.03 (0.77–1.37) | 1.01 (0.77–1.32) | 1.03 (0.41–2.55) | 0.91 (0.33–2.47) | 0.87 (0.60–1.26) |
| Fair to poor general health state (ref: good to excellent) | 0.94 (0.54–1.65) | 1.23 (0.88–1.71) | 1.01 (0.77–1.31) | 1.48 (0.41–5.37) | 1.08 (0.29–4.02) | 1.88 (1.28–2.77) |
| Independent with ADL (no) (ref: yes) | 1.46 (0.42–5.09) | 0.83 (0.48–1.41) | 0.52 (0.31–0.87) | 258.82 (†) | 184.87 (†) | 2.65 (1.54–4.55) |
| Depression (yes) (ref: none) | 1.29 (0.68–2.44) | 0.77 (0.48–1.23) | 0.70 (0.45–1.08) | 1.30 (0.36–4.75) | 0.92 (0.20–4.26) | 1.22 (0.64–2.33) |
| Heart disease present (ref: absent) | 1.39 (0.90–2.13) | 1.78 (1.32–2.39) | 1.38 (1.06–1.78) | 5.53 (1.25–23.61) | 5.25 (1.17–23.62) | 1.56 (1.09–2.22) |
| Hypertension present (ref: absent) | 1.45 (0.99–2.11) | 1.17 (0.90–1.52) | 1.20 (0.95–1.51) | 1.73 (0.76–3.90) | 1.59 (0.67–3.78) | — |
| Treatment of diabetes (ref: no pharmacologic therapy) | ||||||
| Oral hypoglycemics only | 1.77 (1.14–2.76) | 1.34 (0.98–1.83) | 1.38 (1.05–1.81) | 3.67 (1.52–8.81) | 4.81 (1.81–12.81) | 1.37 (0.89–2.11) |
| Insulin only | 1.12 (0.69–1.84) | 1.53 (1.05–2.22) | 1.86 (1.33–2.60) | 4.47 (1.26–15.90) | 7.26 (1.60–33.00) | 1.86 (1.14–3.04) |
| Both oral hypoglycemics and insulin | 1.88 (0.90–3.92) | 1.12 (0.68–1.83) | 2.20 (1.41–3.45) | 6.53 (0.83–51.67) | 6.80 (0.85–54.21) | 0.73 (0.33–1.62) |
| Country (Canada) (ref: U.S.) | 4.01 (2.26–7.14) | 0.33 (0.24–0.46) | 1.46 (1.08–1.98) | 3.02 (1.03–8.84) | 3.35 (1.01–12.81) | 1.79 (1.17–2.75) |
Data are OR (95% CI).
Medical specialist category excludes eye specialists;
extremely large CI including 1.00. Boldface data indicate significance.
Article Information
This work was supported by ACHORD at the Institute of Health Economics, Edmonton, AB, Canada. S.W.K. is supported by a grant from the Alberta Heritage Foundation for Medical Research
References
Address correspondence and reprint requests to Scott Klarenbach, Institute of Health Economics, #1200, 10405 Jasper Ave., Edmonton, Alberta, T5J 3N4 Canada. E-mail: [email protected].
Received for publication 2 October 2002 and accepted in revised form 10 January 2003.
A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.
