Higher quality primary care is associated with good self-rated health status

Transcription

1 Family Practice 2013; 30: doi: /fampra/cmt021 Advance Access publication 12 June 2013 The Author Published by Oxford University Press. All rights reserved. For permissions, please Higher quality primary care is associated with good self-rated health status Nak Jin Sung a, Jeffrey F Markuns b, Ki Heum Park a, Kyoungwoo Kim c, Heeyoung Lee d and Jae Ho Lee e, * a Department of Family Medicine, Ilsan Hospital, Dongguk University College of Medicine, Goyang, Republic of Korea, b Department of Family Medicine, BUFM Global Health Collaborative, Boston University Medical Center, Boston, MA 02118, USA, c Department of Family Medicine, Seoul Paik Hospital, Inje University College of Medicine, Seoul, Republic of Korea, d Cancer Control Center, Gachon University Gil Hospital, Incheon, Republic of Korea and e Department of Family Medicine, The Catholic University of Korea, School of Medicine, Seoul, Republic of Korea. *Correspondence to Jae Ho Lee, Department of Family Medicine, The Catholic University of Korea, School of Medicine, 505 Banpo-dong, Seocho-gu, Seoul , Republic of Korea; jaeholee@catholic.ac.kr Received 2 October 2012; Revised 20 April 2013; Accepted 25 April Objective. To ascertain the association between primary care quality and self-rated health status. Methods. A cross-sectional study using the Korean primary care assessment tool (K-PCAT). The K-PCAT is a validated tool based on the definition of primary care in Korea, consisting of 5 domains and 21 items providing a total primary care quality score. Data were collected from patients of family physicians working at nine private clinics as their usual source of care. The main outcome measure was self-rated health status. Results. Data were analyzed for 531 study participants. Bivariate analysis of socio-demographic variables of patients, who participated in this study as primary care quality assessors, revealed that those with high self-ratings of health tended to have higher household incomes and more frequent exercise. Those with high self-ratings of health had higher total primary care scores than those with low self-ratings of health, as determined through bivariate analysis (P < 0.01). After being adjusted for age, sex, the number of diseases being treated, education years, household income, smoking status, alcohol intake and the frequency of regular exercise, the total primary care score was found to be positively associated with good health. Conclusions. Primary care quality, as assessed by the K-PCAT, was positively associated with good self-rated health status. Keywords. Continuity of care, doctor patient relationship, health promotion, primary care, quality of care. Introduction Primary care has been shown to be associated with a wide range of benefits such as much lower under-five mortality rates, decreased mortality, lower infant mortality, lower incidence of low birth weight, decreased inpatient admission, decreased outpatient visits, decreased emergency room visits and generally better health at lower costs and decreased hospitalizations for primary care sensitive conditions Starfield et al. 5 provided evidence of the benefits of comprehensive primary care by way of international comparisons, studies within countries and studies of the impact of important features of primary care. Starfield 10 reported that the beneficial effects of primary care were evident not only in industrialized countries but also in middle and lower income countries. Among the Organization for Economic Co-operation and Development (OECD) countries, Korea s health system is one of the weakest in terms of primary care orientation. 11 The health care system has been dominated by the private sector, representing >90% of all existing health care facilities including both outpatient facilities and hospitals. 12 Primary care has largely been undervalued by the government, as well as providers and patients, for over 30 years. 13 Medical specialists who run their own community clinics see patients directly, without referrals from family physicians, as there is no gatekeeping function for primary care in Korea. Private sector dominance in the health care system has led South Korea to exhibit the characteristics of a free market, despite a nationalized universal health care program. 14 The fee schedule for primary care providers is based on fee-for-service (rather than on capitation), resulting in physicians lacking an 568

2 Primary care and self-rated health 569 incentive to focus on improving health prevention and promotion. 12 Under these circumstances, induced demand by physicians and medical shopping by patients are more likely to take place, consistent with the finding that South Korea has a relatively high level of outpatient contacts per year per person compared to other OECD countries (12.9 versus 6.4 times in 2010). 15 The government of South Korea attempted to introduce a primary care gatekeeping scheme in 1996, after considering the benefits of primary care. This effort failed, however, due to the resistance of the Korean Medical Association, lack of consensus in public opinion, insufficient government commitment and scant evidence for the effectiveness of such a system in South Korea. Stronger local evidence is needed to substantiate the value of primary care for policy makers in Korea. Self-rated health status is a composite measure to assess physical and psychological health status of individuals. Previous studies have shown that poor assessments of self-rated health had a close association with an increased rate of mortality, increased prevalence of chronic diseases and increased number of visits to clinics. 16,17 Self-rated status can, therefore, be used as a potential surrogate variable for overall health outcome. Meanwhile, there have only been a few studies analyzing the association between primary care quality and self-rated health status as a health outcome, even though high primary care quality is associated with good health indicators. 8,18,19 Furthermore, few studies have investigated the association between primary care quality and self-rated health status at an individual level. We sought to determine if there was a positive association between good primary care experience and health outcomes of patients in Korea. Methods This study was approved by the Institutional Review Board of the Catholic University of Korea. Participant practices We used a part of previous study data that were originally collected to evaluate the validity of the Korean primary care assessment tool (K-PCAT). 20 We confined study subjects to primary care community practices and to those under the care of family physicians, since family medicine is the only specialty with an objective of educating qualified primary care physicians in South Korea. Family physicians have a uniformly similar type of training due to the standardized national residency curriculum, with family medicine having been introduced as a formal specialty in South Korea in In addition, family practices share important common features in that they are remunerated by the fee-for-service method under the national health insurance system, and almost all are privately owned and run by a single doctor (93.5% in 2008). 21 Of community private clinics, we excluded from our study various specialty clinics such as obesity, pain and anti-aging clinics. Only providers who had been practicing in the same clinic for at least 2 years were included in this study. Nine representative Korean primary care practices, similar to the majority of primary care practices in Korea, were selected by the authors using purposive sampling methods. Of the nine participating practices, three were located in Seoul, three in satellite suburban cities and three in rural areas. The nine participating family physicians had practiced family medicine for an average of 16 years. Survey data by the Korean Medical Association (2005) showed that 83.4% of all family physicians worked at similar community private clinics. They provided consultation to ~64 patients in a 10-hour day and saw patients five and a half days per week. 22 Most of the patient visits were for acute or chronic care, rather than prevention. Drop-in visits made without a prior appointment constitute a significant proportion of these consultations. Patient selection and the definition of the usual source of care Surveyed patients consisted of people who visited one of the participating practices and agreed to complete the questionnaire before seeing their physicians. Eligible participants were individuals for whom the practice served as their usual source of care. The usual source of care was defined as a provider whom the user had visited at least six times over a period of >6 months. We used the above definition related to continuity of care from the K-PCAT developmental stage instead of three questions from Starfield s PCAT in identifying the usual source of care and the strength of that affiliation. 21 The percentage of patients who have a usual source of care is much lower in South Korea compared with the USA due to the different health care systems. 23 For this study, we felt it necessary to use this continuity of care measure as an important part of the selection criteria for participants, because patients with limited continuity may have difficulty assessing their experience with primary care Data collection Interviewers were trained in a standardized technique for survey administration. Three interviewers visited each family practice, administered questionnaires to study subjects and helped them answer the questionnaires. The questionnaires were answered by patients themselves. The questionnaires included K-PCAT items and those of general patient characteristics, including

3 570 Family Practice The International Journal for Research in Primary Care a self-rating of health. Data collection was performed from 23 April 2007 to 23 June K-PCAT The K-PCAT is a validated tool developed under the conceptual framework of a Korean definition of primary care. 13 While the late Barbara Starfield of Johns Hopkins University provided helpful comments during its development process, the Korean developers chose to significantly alter the original tool by removing some core primary care elements in an effort to better reflect the current Korean concept of primary care. As such, the tool is designed only for use within the Korean context and is not intended for use between individual countries or for comparison with data generated through other PCAT tools that are intended to more closely maintain integrity with the original tool. The K-PCAT consists of four multi-item scales and one composite scale (Appendix). 21 The scoring system is structured as follows: each response on a five-point Likert scale is converted from 0 to 4. The means of item scores in the same domain are multiplied by 25 to yield domain scores (0 100). The primary care average score is the mean of five domain scores and believed to reflect an overall measure of the quality of primary care by reflecting the degree of practice adherence to several core primary care principles. Statistical analysis As a measure of overall health, the respondents selfrated health status was used and coded on a binary scale as 1 for respondents reporting very good or good health (herein referred to as good health) and 0 for those reporting neutral, poor or very poor health (herein referred to as poor health), following the method of previous studies for comparability. 18,24 The socio-demographic data of the participating patients by self-rated health status were analyzed by the Mantel Haenszel chi-square test. Attribute scores of primary care on the K-PCAT were analyzed by Student s t-test. In order to control for age, sex, the number of diseases being treated, education years, income and health behaviours, we used multiple logistic regression analysis. We assessed the minimum number of cases of 500 by the rule of thumb formula in the logistic regression analysis. 25 Let p be the smallest of the proportions of negative or positive cases in the population and k the number of covariates, then the minimum number of cases to include is N = 10k/p. The maximum number of covariates equals 15 in Model V of Table 3. The proportion of good health in our patient population was assessed to be 0.3 because the proportion of good health was 0.32 in the 2007 Korean Health and Nutrition Examination Survey. 26 The statistical software STATA SE 11 was used for the analysis. Results Of those who visited their family practice as a usual source of care and thus were eligible for the previous study (n = 726), there were no significant differences between the participants (n = 608, 83.7%) and nonparticipants (n = 118, 16.3%) in terms of age and sex. The most common reason for refusing to complete the questionnaire was that the patient was too busy. Sixty-nine patients aged 17 years or less were originally surveyed but later excluded from our analysis due to their duration of education, typically being prior to high school graduation. The decision to exclude minors was made because duration of education was an important factor closely related to self-rated health status, the dependent variable, in this study. The duration of education was <12 years for the age group of 17 or less. Another six patients were excluded for failing to complete more than three items, and two more were excluded for failing to complete the self-rating of health. In sum total after exclusions, data for 531 study participants were used for all analyses (Fig. 1). Distributions of socio-demographic variables by selfrated health status Bivariate analysis of socio-demographic variables of patients who participated in this study revealed that the good health group tended to be older with a longer duration of education, smaller number of diseases being treated, greater household income and more frequent regular exercise than the poor health group (P < 0.05). There was no significant difference in overall health status between genders (Table 1). Previous study subjects (726) a 608 Eligible subjects for this study (539) Subjects for analysis (531) Excluded for refusing to participate (118) Excluded for age of 17 or less (69) b Excluded for missing on more than 3 items (6) and on self-rated health item (2) Figure 1 Data selection process. a We used a part of data, which were originally collected for K-PCAT validation study. b The original data included all age group. In this study, we excluded data for age of 17 or less who were prior to high school graduation because duration of education was an important factor closely related to self-rated health status, the dependent variable in this study

4 Primary care and self-rated health 571 Table 1 Socio-demographic variables by self-rated health: frequency (%) Self-rated health Total (n = 531) Poor health (n = 413) Good health (n = 118) P-value a Age (years) (20.5) 69 (16.7) 40 (33.9) < (50.7) 212 (51.3) 57 (48.3) 65 or more 153 (28.8) 132 (32.0) 21 (17.8) Sex Male 188 (35.4) 146 (35.4) 42 (35.6) 0.96 Female 343 (64.6) 267 (64.6) 76 (64.4) Education (years) (39.0) 173 (43.1) 28 (24.6) < (30.7) 118 (29.4) 40 (35.1) > (30.3) 110 (27.4) 46 (40.4) Missing (16) Number of diseases of being treated (29.9) 111 (26.9) 48 (40.7) < (47.7) 195 (47.2) 58 (49.2) 2 or more 119 (22.4) 107 (25.9) 12 (10.2) Income (million Won b /month/ household) (33.3) 140 (35.8) 26 (24.3) < (29.3) 122 (31.2) 24 (22.4) (37.4) 129 (33.0) 57 (53.3) Missing (33) Alcohol intake No 360 (67.8) 296 (71.7) 64 (54.2) <0.01 Moderate 126 (23.7) 85 (20.6) 41 (34.8) Excessive 45 (8.5) 117 (7.7) 13 (11.0) Regular exercise (frequency/week) (34.4) 160 (38.9) 22 (18.6) < (25.0) 96 (23.4) 36 (30.5) 3 or more 215 (40.6) 155 (37.7) 60 (50.9) Smoking status No smoker 354 (66.7) 275 (66.6) 79 (67.0) 0.97 Ex-smoker 74 (13.9) 57 (13.8) 17 (14.4) Current smoker 103 (19.4) 81 (19.6) 22 (18.6) Poor health: people with high self-ratings of health (neutral, poor or very poor). Good health: people with low self-ratings of health (very good or good). a P-value by chi-square test. b One US dollar equals 956 Won on the day of survey (1 April 2007). Average household income was Won per month during the second quarter of 2007 in South Korea. Primary care quality (attribute scores) of the five domains of the K-PCAT analyzed by self-rated health status The bivariate analysis test for the five domains of primary care attributes revealed that the family/ community orientation domain score and total primary care score were significantly higher in the good health group than in the poor health group (P < 0.01). In the other domains, primary care scores were not different between the two groups (Table 2). Factors associated with good health Five models were utilized to assess relationships of primary care scores and socio-demographic factors with self-rated health status. All models were adjusted for age, sex and number of diseases being treated. Model I assessed the total primary care score without further adjustment. Model II analyzed domain scores without further adjustment and Model III analyzed domain scores adjusted for one another. Model IV included the total primary care score adjusted for education and income, with Model V including additional adjustment for smoking, alcohol use and exercise (Table 3). The total primary care score was positively associated with good health after both partial and full adjustment for socio-demographic variables (Models I, IV and V; P < 0.05). After adjusting for age, sex and the number of diseases being treated, the domain scores of first contact, personalized care and family/community orientation were all associated with good health (Model II). Only family/community orientation scores, however, remained associated with good health after adjusting for all domain scores (Model III; P < 0.05). In Models IV and V, in terms of household income,

5 572 Family Practice The International Journal for Research in Primary Care Table 2 Patient assessment of quality (attribute scores) of primary care by self-rated health, using the K-PCAT: mean ± SD Domains No. of items Total (n = 524) Self-rated health Poor health a (n = 408) Good health b (n = 116) P-value c Personalized care ± ± ± Coordination function ± ± ± Comprehensiveness ± ± ± Family/community orientation ± ± ± 18.2 <0.01 First contact d ± ± ± Total primary care score e ± ± ± 15.9 <0.01 a Poor health: people with high self-ratings of health (neutral, poor or very poor). b Good health: people with low self-ratings of health (very good or good). c P-value by t-test. d First contact, composite domain, consists of score average of five independent subscales. e Score average of five domains. Table 3 Factors associated with good health a compared to poor health b Odds ratio (95% confidence interval) c Model I Model II d Model III d Model IV Model V Total primary care score e 1.02 ( ) f 1.02 ( ) f 1.02 ( ) f Primary care domain score First contact score 1.02 ( ) f 1.01 ( ) Personalized care score 1.03 ( ) f 1.01 ( ) Coordination function score 1.00 ( ) 0.99 ( ) Comprehensiveness score 1.01 ( ) 1.00 ( ) Family/community orientation 1.03 ( ) f 1.02 ( ) f score Education (years) ( ) 1.18 ( ) ( ) 1.18 ( ) > Income (million Won g /month/household) ( ) 0.70 ( ) ( ) f 0.50 ( ) f Smoking status Non-smoker 1 Ex-smoker 0.82 ( ) Current smoker 0.93 ( ) Alcohol intake No 1 Yes 1.80 ( ) f Regular exercise (frequency/week) ( ) 3 or more 2.55 ( ) f a Good health: people with low self-ratings of health (very good or good). b Poor health: people with high self-ratings of health (neutral, poor or very poor). c Odds ratio and 95% confidence interval were calculated by multiple logistic regression analysis. d Each domain score was included individually in Model II and all together in Model III using multiple logistic regression. e Score average of five domains. f P-value < 0.05 by multiple logistic regression analysis (n = 600). g One US dollar equals 956 Won on the day of survey (1 April 2007). Average household income was Won per month during the second quarter of 2007 in South Korea.

6 Primary care and self-rated health 573 the group with a monthly income of million Won was negatively associated with good health compared to the group earning 3.0 million Won or more (P < 0.05). However, the odds ratio of good health for the group with a monthly income of 1.4 million Won or less was not significantly different from that of the group earning 3.0 million Won or more (P > 0.05). Discussion This study aimed to determine if there is an association between quality of primary care and health outcomes. Specifically, we sought to ascertain whether a score of primary care attributes as reported by patients, intended to represent primary care quality, would be associated with self-rated health status, intended as a proxy for overall health outcomes. Without any adjustments, the total primary care score was positively associated with good self-rated health status (Table 2). This association held true with both partial and full adjustment for socio-demographic factors (Table 3), indicating the presence of a robust association. This finding is consistent with a prior study showing that a good experience with primary care, including satisfaction with care, may be associated with self-rated health status. 27 To our knowledge, this is the first study to examine the association between a total primary care score consisting of primary care attribute domain scores and selfrated health status. There have been only two related studies in the past, and both were focused on health care in the USA. Using the 1996 Community Tracking Study (CTS) household survey, Shi and Starfield 18 examined whether income inequality and primary care measured at the state level predicted individual morbidity, as measured by self-rated health status, while adjusting for potentially confounding individual variables. Their results indicated that individuals living in states with higher ratios of primary care physicians are more likely to report good health than those living in states with lower ratios of primary care physicians. Using the CTS household survey and state indicators of primary care, a subsequent study by Shi et al. 24 showed that good primary care experience was also associated with improved self-rated overall health after adjusting for socioeconomic factors, primary care doctor population ratio and usual source of care. The primary care experience in this study was calculated by the summation of item scores from the CTS deemed to represent accessibility, interpersonal relationship and continuity of care. However, item scores for comprehensiveness, coordination and accountability domains were not included because there were no suitable questionnaire items in the CTS household survey. In Korea, we have focused on the K-PCAT as a way of determining high-quality primary care and assessing elements of primary care practice. One challenging domain to assess in Korea, however, is continuity of care. In Korea, a primary care physician may or may not act as a usual source of care to patients, making the assessment of continuity of care somewhat more difficult. 21 Diabetics receiving care from a single attending clinic for a long period of time, however, showed lower admission rates and lower complication rates in a Korean study. 28 In a further Korean study, highquality primary care was associated with good health behaviours. 29 These studies have similarities to our study in that better primary care experience was associated with better outcomes or healthier behaviours. Due to the cross-sectional nature of this study, it is difficult to identify a causal relationship between primary care experience and self-rated health. Certainly, one can envision how good quality primary care represented by high K-PCAT scores could result in improved health status and thus good self-rated health. Alternatively, those with good self-ratings of health may have more positive experiences with primary care, thus resulting in higher quality ratings of primary care and high K-PCAT scores. Indeed, previous studies have shown that patients with good health status were more likely to be satisfied with health care and a new form of out-of-hours care. 27,30,31 In the current study, those with good self-ratings of health had fewer chronic diseases and were younger in age. It is unclear if such characteristics lead patients to rate their primary care physicians more highly, or if having fewer diseases may simply be a result of accessing higher quality primary care. Of the five primary care domain scores on the K-PCAT, domain scores for first contact, personalized care and family/community orientation were all significantly associated with good health (Model II), while only the family/community orientation domain score showed a significant association after adjusting for all domain scores (Model III). Consistent with the idea that these domains together represent some of the most important quality features of primary care, these findings suggest the domain scores are closely associated with one another and thus influence self-rated health in a collaborative fashion. Determining why some domain scores may be associated with good health and others are not requires further study. Analysis of the association between household income and self-rated health status also yielded interesting results (Table 3). The middle income group showed significantly lower odds of good health than the upper income group, while the lower income group appeared to have similar reported health status as the upper income group. These findings were different from a similar US-based study, which showed a consistent positive association between family income and selfrated health status. 24 This difference may partially be explained by the Korean Medical Aid Program. Insured

7 574 Family Practice The International Journal for Research in Primary Care individuals in South Korea, even those with national health insurance, are required to pay a certain portion of their health care costs, and these co-payments can be burdensome to the middle income group. 32 Meanwhile, those in the low-income group qualify for additional government aid that helps to greatly decrease their economic burden from health care utilization, therefore potentially resulting in better self-rated health status. Our study had several limitations. First, the scores were based on patient assessment, which reflected their experiences rather than specific objective measures of primary care quality. Patient perceptions are under the influence of many factors and often do not represent the whole picture. This method is highly patient centred, however, being derived directly from the patient experience. Second, there could be a selection bias, particularly given accessibility is an important component of the first contact domain in the K-PCAT. The selection bias might be inevitable, but not significant in this study, as the scores of the first domain including accessibility items were shown to be much higher than those of other domains by previous studies using K-PCAT. 21,33 This may result from diminished barriers to access as a result of national health insurance. Third, there were unadjusted and remaining confounding variables, such as income inequality, which might have influenced selfrated health status and may merit future study. Conclusions Primary care quality, as measured in Korea using a total primary care score as well as scores in specific primary care domains, is associated with better self-rated health status. This association holds even when adjusted for a variety of socioeconomic and demographic variables. This study suggests that the provision of high-quality primary care incorporating key principles, such as first contact care, personalized care and family and community orientation, may result in improved health outcomes. Declaration Funding: none. Ethical approval: the Institutional Review Board of the Catholic University of Korea. Conflict of interest: none. References 1 Starfield B. Primary Care: Concept, Evaluation, and Policy. London: Oxford University Press, Shi L. The relationship between primary care and life chances. J Health Care Poor Underserved 1992; 3: Starfield B. Primary care: is it essential? Lancet 1994; 344: Bindman AB, Grumbach L, Osmond D. Primary care and receipt of preventive services. J Gen Intern Med 1996; 11: Starfield B, Shi L, Macinko J. Contribution of primary care to health systems and health. Milbank Q 2005; 83: Rohde J, Cousens S, Chopra M et al. 30 years after Alma-Ata: has primary health care worked in countries? Lancet 2008; 372: Kruk ME, Porignon D, Rockers PC, Van Lerberghe W. The contribution of primary care to health and health systems in low- and middle-income countries: a critical review of major primary care initiatives. Soc Sci Med 2010; 70: Kringos DS, Boerma WG, Hutchinson A, van der Zee J, Groenewegen PP. The breadth of primary care: a systematic literature review of its core dimensions. BMC Health Serv Res 2010; 10: Friedberg MW, Hussey PS, Schneider EC. Primary care: a critical review of the evidence on quality and costs of health care. Health Aff 2010; 29: Starfield B. Primary care: an increasingly important contributor to effectiveness, equity, and efficiency of health services. SESPAS report Gac Sanit 2012; 26 (suppl 1): Ahn SH. Assessment of primary care level in Korea and comparison with the developed countries. J Korean Acad Fam Med 2001; 22: Chun CB, Kim SY, Lee JY, Lee SY. Republic of Korea: health system review. Health Syst Trans 2009; 11: Lee JH, Choi YJ, Volk RJ et al. Defining the concept of primary care in South Korea using a Delphi method. Fam Med 2007; 39: Peabody JW, Lee SW, Bickel SR. Health for all in the Republic of Korea: one country s experience with implementing universal health care. Health Policy 1995; 31: Organization for Economic Co-operation and Development. OECD Health Data 2012 Frequently Requested Data. Paris, France: OECD, (accessed on 16 May 2013). 16 Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997; 38: Jee SH, Ohr HC, Kim IL. Self-rated health and mortality in elderly. Epidemiol Health 1994; 16: (in Korean). 18 Shi L, Starfield B. Primary care, income inequality, and self-rated health in the United States: a mixed-level analysis. Int J Health Serv 2000; 30: Gravelle H, Morris S, Sutton M. Are family physicians good for you? Endogenous doctor supply and individual health. Health Serv Res 2008; 43: Lee JH, Choi YJ, Sung NJ et al. Development of the Korean primary care assessment tool-measuring user experience: tests of data quality and measurement performance. Int J Qual Health Care 2009; 21: Im GJ, Min HY, Choi JW, Lim SM, Park YH. Financial state of primary care physicians under the Korean insurance system. J Korean Med Assoc 2011; 54: Lim GJ. Business analysis of the practitioners offices in Korea Research report Research Institute for Healthcare Policy, Jung YH. A report on the Korea Health Panel Survey: health care utilizations and out-of-pocket spending. Health Welfare Policy Forum 2011; 179: Shi L, Starfield B, Politzer R, Regan J. Primary care, self-rated health, and reductions in social disparities in health. Health Serv Res 2002; 37: Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol 1996; 49: Korean Health and Nutrition Examination Survey (KHANES) report. (accessed on 18 December 2012). 27 Ren XS, Kazis L, Lee A, Rogers W, Pendergrass S. Health status and satisfaction with health care: a longitudinal study among patients served by the veterans health administration. Am J Med Qual 2001; 16: Kim JY, Kim HY, Kim HY et al. Current status of the continuity of ambulatory diabetes care and its impact on health outcomes and medical cost in Korea using National Health Insurance Database. J Korean Diabetes Assoc 2006; 30:

8 Primary care and self-rated health Park JH, Kim KW, Sung NJ et al. Association between primary care quality and health behaviors in patients with essential hypertension who visit a family physician as a usual source of care. Korean J Fam Med 2011; 32: Zhang Y, Rohrer J, Borders T, Farrell T. Patient satisfaction, selfrated health status, and health confidence: an assessment of the utility of single-item questions. Am J Med Qual 2007; 22: Glynn LG, Byrne M, Newell J, Murphy AW. The effect of health status on patients satisfaction with out-of-hours care provided by a family doctor co-operative. Fam Pract 2002; 21: Song YJ. The South Korean health care system. JMAJ 2009; 52: Sung NJ, Suh SY, Lee DW et al. Patient s assessment of primary care of medical institutions in South Korea by structural type. Int J Qual Health Care 2010; 22: Appendix. Korean Primary Care Assessment Tool (K-PCAT) A nine expert panel developed the preliminary 30-item tool. Five items were eliminated after discussion with a practicing physician expert panel, and another four items were eliminated by principle component analysis, leaving 21 total items. Primary care attributes First contact Personalized care Coordination function Comprehensiveness Family/community orientation Item contents Do you visit this clinic first when a new health problem arises? Is it easy for you to access this facility? Appropriateness of out-of-pocket cost? Your doctor sees patients regardless of their age and sex? Basic health care available? Does your doctor treat mental health problems as well as physical health problems? Doctor understands patients words easily? Doctor explains test results in a manner that is easy for patients to understand? Doctor recognizes the importance of patients medical history? Do you trust your doctor s decisions on treatment? Does your doctor recommend health care resources appropriately? Since your doctor started treating you, have you ever visited a specialist? If yes, did your doctor recommend the specialist? Since your doctor started treating you, have you ever visited a specialist? If yes, did your doctor review the referral results? Medical check-ups available? (e.g. physical exam, blood sugar, cholesterol, blood pressure controls, etc.) Does your doctor counsels for cancer prevention and screening? Do you (or your family member) get periodic Pap smear tests from your physician? Periodic health examination by your physician? Doctor knows about the health, well-being and environmental problems of your community? Doctor has a concern about the persons living with you? Is the doctor active in promoting the health of your community? Does this clinic surveys and reflects people s opinions on health care?