https://orcid.org/0000-0002-8232-2155
Abstract
The existence of a stable usual source of care (USC) is fundamental to the provision of quality health care. However, no longitudinal studies have examined whether core primary care attributes influence the stability of USC status.
We aimed to examine the association between primary care attributes (first contact, longitudinality, coordination, comprehensiveness, and community orientation) and the loss or change of USC.
This nationwide cohort study was conducted during the coronavirus disease 2019 pandemic using a representative sample of the Japanese adult population aged 40–75 years. The primary outcome measures were loss of USC and voluntary change in USC during the 12-month follow-up period. Primary care attributes were evaluated in the baseline survey using the Japanese version of Primary Care Assessment Tool (JPCAT).
Data were analyzed for 725 participants who had a USC at baseline. Among them, 93 (12.8 %) lost their USC and 46 (6.3%) changed their USC during the follow-up period. Multivariable multinominal logistic regression analyses showed that the JPCAT total score was associated with decreased loss of USC and change in USC. Among the JPCAT domains, longitudinality, comprehensiveness (services available), and community orientation were associated with reductions in both USC loss and change.
Our study indicates that primary care attributes play an important role in preventing the loss or change of USC and contribute to the stability of USC status. These findings provide additional rationale for policymakers, healthcare providers, and managers to seek to strengthen core attributes of primary care.
Primary care attributes were associated with reductions in USC loss and change.
Key attributes were longitudinality, comprehensiveness, and community orientation.
Improving primary care attributes can help increase residents with stable USC.
Introduction
A usual source of care (USC) is a person or place that people can turn to with a health issue or concern.1 According to previous studies, having a USC has been linked to increased preventive services,2,3 better control of chronic diseases,4 fewer emergency department visits,5 fewer total hospitalizations,6 and lower healthcare costs.7 The loss of a USC can be a significant negative event for individuals because they will not be able to receive these healthcare benefits. In addition, individuals who have changed USCs also have as many unmet healthcare needs as those who lacked a USC, compared with those with a continuous USC relationship.8,9 Therefore, the existence of a stable USC is fundamental to the provision of quality health care.
A wealth of cross-sectional studies of the presence or absence of USC at a single point in time have been reported internationally. For example, in exploratory cross-sectional studies conducted in the USA, Korea, and Japan, various resident factors, including age, gender, marital status, region, education, race, income, insurance, chronic conditions, health literacy, and social network, have been found to be associated with the presence of USC.1,10–12 In addition, a cross-sectional study conducted in China demonstrated the association between primary care quality as a provider factor and the presence of USC.13
On the other hand, longitudinal studies of the stability of USC status are very limited. An exploratory longitudinal study conducted in the USA reported that the odds of losing a USC were higher among older adults who had unmet transportation needs, moved to a new residence, reported depressive symptoms, had fewer chronic conditions, and were uninsured.14 Another practice-level longitudinal study from the UK found that primary care practices with low levels of patient satisfaction were more likely to have high rates of changing practices.15 However, no longitudinal studies have examined whether core primary care attributes (first contact, longitudinality, coordination, comprehensiveness, and community orientation) influence the loss or change of USC. First-contact care means that care is first sought from the primary care provider when a new health or medical need arises. Longitudinality refers to the longitudinal use of a regular Source of care over time, regardless of the presence or absence of disease or injury. Coordinated care is the linking of health care visits and services so that patients receive appropriate care for all of their health problems. Comprehensive care refers to the availability of a wide range of services in primary care and their appropriate provision across the entire spectrum of types of needs. Community-oriented care refers to care that is delivered in the context of the community.16
In Japan, physicians who finish their internal medicine training and subsequently practice in clinics or smaller hospitals with less than 200 beds frequently function as USCs, called kakaritsukei in Japanese.17 There are differences in beds, inpatient care, and staff size between clinic-based and hospital-based practices.18,19 Since 2010, the Japan Primary Care Association has begun certifying family physicians, and its training system received international certification from the World Organization of Family Doctors in 2020.20 Meanwhile, a new board certification system for primary care specialists by the Japanese Medical Specialty Board, which is independent of existing specialty bodies, began in 2018.18 But at present, the number of such specialists is limited. There is no registration system for primary care physicians in Japan, and residents are free to choose and change their own USC.18 Patients have direct access to large hospitals, including university hospitals, but few residents use large hospitals as USCs because of the additional costs.17 Although the Ministry of Health, Labour and Welfare has recommended that all individuals should have a USC, the proportion of adult residents who had a USC was 57.5% in 2021.12 Because some primary care facilities, including both clinics and hospitals in Japan, did not accept patients with coronavirus disease 2019 (COVID-19),21 it was assumed that USC change and loss were likely to occur and that the pandemic was a suitable setting for a study using these as outcomes.
In the present study, we aimed to examine the association between core primary care attributes and the loss or change of USC based on data from a nationwide cohort study.
Methods
Design, setting, and participants
We used the data from the National Usual Source of Care Survey, which is a nationwide cohort study conducted between May 2021 and April 2022 during the COVID-19 pandemic (follow-up period: 12 months). The National Usual Source of Care Survey was a nationwide mail survey designed to assess the USC, the quality of primary care, health care utilization, health status, and sociodemographic factors in a representative sample of Japanese adults. In this study, a nationally representative panel, administered by the Nippon Research Center, was used to select potential participants. This panel comprised approximately 70 000 residents who were selected from the Japanese adult population using a multistage sampling method. From the panel, prospective participants aged 40–75 years were selected through stratified random sampling by age, sex, and residential area.6 Several research questions were included in the National Usual Source of Care Survey, and the overall sample size was determined by the needs of other research questions that are part of the National Usual Source of Care Survey.3,6
Among National Usual Source of Care Survey respondents, individuals who had a USC at baseline were eligible for this study. To identify an individual’s USC, we applied a specific item used in the Japanese version of Primary Care Assessment Tool (JPCAT)22,23 and previous studies1,14: “Is there a physician that you usually go to if you are sick or need advice about your health?” A participant was considered to have a USC if he or she could identify a physician who practiced outside of university hospitals. This is because the Ministry of Health, Labour and Welfare of Japan separates the roles of kakaritsukei, which provide primary care, and university hospitals.
The institutional review board of the Jikei University School of Medicine approved this study (approval no. 32-416(10505)).
Measures
Primary care attributes were assessed in the baseline survey using the JPCAT, which is based on the PCAT, a well-known and globally accepted scale for assessing primary care attributes.24,25 Previous studies have demonstrated the reliability and validity of the JPCAT.22,23 It consists of six domains, each containing multiple items that address core attributes of primary care: first contact, longitudinality, coordination, comprehensiveness (services available), comprehensiveness (services provided), and community orientation. As a difference from the original PCAT, community orientation was identified as one of the core primary care attributes in Japan during the JPCAT development process.22 In this study, we used the 13-item abbreviated version of the scale.23 The JPCAT scoring system is based on a 5-point Likert scale (1 = strongly disagree, 2 = somewhat disagree, 3 = not sure, 4 = somewhat agree, and 5 = strongly agree). Each response is converted into an item score ranging from 0 to 4. Item scores within each domain are averaged and multiplied by 25, resulting in domain scores ranging from 0 to 100 points. The JPCAT total score is calculated as the mean of the six domain scores and provides an overall measure of primary care attributes. Higher scores indicate higher quality.
The primary outcome measures in this study were loss of USC and voluntary change in USC during the 12-month follow-up period. Loss of a USC was defined as the transition from reporting having a USC at the baseline survey to reporting no USC at the follow-up survey.14 The item and algorithm used to identify an individual’s USC are described above. Change in USC was measured by the response to the follow-up survey question “Have you voluntarily changed your usual source of care (kakaritsukei) in the past 12 months?” for participants who had a USC at both the baseline and follow-up surveys. Participants were asked on a binary scale (“yes” or “no”).
The following potential confounders were selected on the basis of prior research8,13,14 and clinical experience: age (continuous), gender, educational level (< university or ≥university), annual household income (< 3.00 million JPY, 3.00–6.99, or ≥7.00), number of chronic conditions (continuous), and health-related quality of life (continuous) assessed by the five-level version of the EuroQol five-dimensional questionnaire.26 We used a validated list of 20 chronic conditions based on previous multimorbidity literature and their relevance to the primary care population.27 All confounders were evaluated at baseline using a self-administered questionnaire.
Statistical analysis
For continuous data, descriptive statistics were presented in the form of means and standard deviations, while frequencies and percentages were used to report categorical data. One-way analysis of variance for continuous data and chi-squared test for categorical data were used to compare participant characteristics by USC status.
We used multivariable multinominal logistic regression analyses to determine whether the JPCAT total score was associated with the loss of USC and the change in USC, in reference to stable USC. The following potential confounders were included in the regression model: age, gender, educational level, annual household income, number of chronic conditions, and the five-level version of the EuroQol five-dimensional questionnaire score. In addition, we performed ancillary exploratory analyses of outcomes in relation to each JPCAT domain score.
For each analysis, we used a two-sided significance level of P < 0.05. No adjustment for multiple comparisons was applied in the exploratory analyses. Multiple imputations with 20 imputations using a fully conditional specification were applied to deal with missing data on independent variables. Statistical analyses were conducted using R, version 4.3.0 (R Foundation for Statistical Computing, Vienna, Austria; www.R-project.org).
Results
Sample characteristics
Figure 1 shows the flowchart of participant selection. Of the 1382 eligible participants selected by stratified random sampling from the nationally representative panel, 1,262 completed the baseline survey. Supplementary Table S1 shows the comparison between the population distribution of the Japanese census data and that of the baseline survey participants. Of these, we included 780 participants who had a USC at baseline. Supplementary Table S2 shows the differences in baseline characteristics between participants with and without USC. A total of 725 participants completed the follow-up survey about the loss and change of USC (follow-up rate: 92.9%).
Participant flow chart.
Table 1 shows the baseline characteristics of the study participants. Among the participants, 93 (12.8 %) lost their USC and 46 (6.3%) changed their USC during the 12 months. We found that participants who lost their USC had fewer chronic conditions, and participants who changed their USC were more likely to be female. In addition, participants with a loss of USC or change in USC had lower JPCAT total and domain scores, except for first contact.
Baseline participant characteristics by usual source of care status.
| Characteristic | Overall | USC status | P-valuea | ||
|---|---|---|---|---|---|
| (n = 725) | Stable USC (n = 586) | Loss of USC (n = 93) | Change in USC (n = 46) | ||
| Age, mean (SD), y | 59.9 (10.1) | 60.1 (10.0) | 58.0 (10.7) | 61.0 (9.1) | 0.143 |
| Gender, n (%) | 0.029 | ||||
| Female | 383 (52.8) | 303 (51.7) | 47 (50.5) | 33 (71.7) | |
| Educational level, n (%) | 0.068 | ||||
| University or higher | 208 (28.7) | 163 (27.8) | 35 (37.6) | 10 (21.7) | |
| Data missing | 16 (2.2) | 12 (2.0) | 4 (4.3) | 0 (0.0) | |
| Annual household income, n (%), million JPY | 0.421 | ||||
| <3.00 (≒ US$21,000) | 143 (19.7) | 119 (20.3) | 19 (20.4) | 5 (10.9) | |
| 3.00–6.99 | 414 (57.1) | 326 (55.6) | 59 (63.4) | 29 (63.0) | |
| ≥7.00 | 160 (22.1) | 135 (23.0) | 14 (15.1) | 11 (23.9) | |
| Data missing | 8 (1.1) | 6 (1.0) | 1 (1.1) | 1 (2.2) | |
| Number of chronic conditions, mean (SD) | 1.6 (1.4) | 1.6 (1.4) | 1.1 (1.4) | 1.9 (1.4) | 0.003 |
| Data missing, n (%) | 22 (3.0) | 16 (2.7) | 5 (5.4) | 1 (2.2) | |
| EQ-5D-5L, mean (SD) | 0.88 (0.09) | 0.88 (0.10) | 0.89 (0.06) | 0.89 (0.06) | 0.216 |
| Data missing, n (%) | 2 (0.3) | 2 (0.3) | 0 (0.0) | 0 (0.0) | |
| JPCAT | |||||
| Total score, mean (SD) | 47.5 (17.0) | 49.3 (16.9) | 39.5 (13.8) | 40.7 (18.1) | <0.001 |
| Domain scores, mean (SD) | |||||
| First contact | 34.3 (26.3) | 35.1 (26.7) | 32.4 (23.6) | 27.5 (26.0) | 0.127 |
| Longitudinality | 57.1 (25.2) | 59.6 (24.6) | 44.9 (23.9) | 49.5 (27.6) | <0.001 |
| Coordination | 56.2 (30.6) | 57.7 (30.9) | 50.8 (26.2) | 48.4 (32.9) | 0.026 |
| Comprehensiveness (services available) | 51.2 (26.3) | 53.4 (26.2) | 42.3 (24.0) | 41.3 (27.1) | <0.001 |
| Comprehensiveness (services provided) | 36.3 (32.0) | 38.5 (32.3) | 23.5 (27.6) | 33.7 (30.7) | <0.001 |
| Community orientation | 50.2 (19.8) | 51.8 (19.3) | 43.0 (18.8) | 44.0 (23.7) | <0.001 |
| Characteristic | Overall | USC status | P-valuea | ||
|---|---|---|---|---|---|
| (n = 725) | Stable USC (n = 586) | Loss of USC (n = 93) | Change in USC (n = 46) | ||
| Age, mean (SD), y | 59.9 (10.1) | 60.1 (10.0) | 58.0 (10.7) | 61.0 (9.1) | 0.143 |
| Gender, n (%) | 0.029 | ||||
| Female | 383 (52.8) | 303 (51.7) | 47 (50.5) | 33 (71.7) | |
| Educational level, n (%) | 0.068 | ||||
| University or higher | 208 (28.7) | 163 (27.8) | 35 (37.6) | 10 (21.7) | |
| Data missing | 16 (2.2) | 12 (2.0) | 4 (4.3) | 0 (0.0) | |
| Annual household income, n (%), million JPY | 0.421 | ||||
| <3.00 (≒ US$21,000) | 143 (19.7) | 119 (20.3) | 19 (20.4) | 5 (10.9) | |
| 3.00–6.99 | 414 (57.1) | 326 (55.6) | 59 (63.4) | 29 (63.0) | |
| ≥7.00 | 160 (22.1) | 135 (23.0) | 14 (15.1) | 11 (23.9) | |
| Data missing | 8 (1.1) | 6 (1.0) | 1 (1.1) | 1 (2.2) | |
| Number of chronic conditions, mean (SD) | 1.6 (1.4) | 1.6 (1.4) | 1.1 (1.4) | 1.9 (1.4) | 0.003 |
| Data missing, n (%) | 22 (3.0) | 16 (2.7) | 5 (5.4) | 1 (2.2) | |
| EQ-5D-5L, mean (SD) | 0.88 (0.09) | 0.88 (0.10) | 0.89 (0.06) | 0.89 (0.06) | 0.216 |
| Data missing, n (%) | 2 (0.3) | 2 (0.3) | 0 (0.0) | 0 (0.0) | |
| JPCAT | |||||
| Total score, mean (SD) | 47.5 (17.0) | 49.3 (16.9) | 39.5 (13.8) | 40.7 (18.1) | <0.001 |
| Domain scores, mean (SD) | |||||
| First contact | 34.3 (26.3) | 35.1 (26.7) | 32.4 (23.6) | 27.5 (26.0) | 0.127 |
| Longitudinality | 57.1 (25.2) | 59.6 (24.6) | 44.9 (23.9) | 49.5 (27.6) | <0.001 |
| Coordination | 56.2 (30.6) | 57.7 (30.9) | 50.8 (26.2) | 48.4 (32.9) | 0.026 |
| Comprehensiveness (services available) | 51.2 (26.3) | 53.4 (26.2) | 42.3 (24.0) | 41.3 (27.1) | <0.001 |
| Comprehensiveness (services provided) | 36.3 (32.0) | 38.5 (32.3) | 23.5 (27.6) | 33.7 (30.7) | <0.001 |
| Community orientation | 50.2 (19.8) | 51.8 (19.3) | 43.0 (18.8) | 44.0 (23.7) | <0.001 |
aP-value by one-way analysis of variance for continuous variables and chi-square test for categorical data.
Baseline participant characteristics by usual source of care status.
| Characteristic | Overall | USC status | P-valuea | ||
|---|---|---|---|---|---|
| (n = 725) | Stable USC (n = 586) | Loss of USC (n = 93) | Change in USC (n = 46) | ||
| Age, mean (SD), y | 59.9 (10.1) | 60.1 (10.0) | 58.0 (10.7) | 61.0 (9.1) | 0.143 |
| Gender, n (%) | 0.029 | ||||
| Female | 383 (52.8) | 303 (51.7) | 47 (50.5) | 33 (71.7) | |
| Educational level, n (%) | 0.068 | ||||
| University or higher | 208 (28.7) | 163 (27.8) | 35 (37.6) | 10 (21.7) | |
| Data missing | 16 (2.2) | 12 (2.0) | 4 (4.3) | 0 (0.0) | |
| Annual household income, n (%), million JPY | 0.421 | ||||
| <3.00 (≒ US$21,000) | 143 (19.7) | 119 (20.3) | 19 (20.4) | 5 (10.9) | |
| 3.00–6.99 | 414 (57.1) | 326 (55.6) | 59 (63.4) | 29 (63.0) | |
| ≥7.00 | 160 (22.1) | 135 (23.0) | 14 (15.1) | 11 (23.9) | |
| Data missing | 8 (1.1) | 6 (1.0) | 1 (1.1) | 1 (2.2) | |
| Number of chronic conditions, mean (SD) | 1.6 (1.4) | 1.6 (1.4) | 1.1 (1.4) | 1.9 (1.4) | 0.003 |
| Data missing, n (%) | 22 (3.0) | 16 (2.7) | 5 (5.4) | 1 (2.2) | |
| EQ-5D-5L, mean (SD) | 0.88 (0.09) | 0.88 (0.10) | 0.89 (0.06) | 0.89 (0.06) | 0.216 |
| Data missing, n (%) | 2 (0.3) | 2 (0.3) | 0 (0.0) | 0 (0.0) | |
| JPCAT | |||||
| Total score, mean (SD) | 47.5 (17.0) | 49.3 (16.9) | 39.5 (13.8) | 40.7 (18.1) | <0.001 |
| Domain scores, mean (SD) | |||||
| First contact | 34.3 (26.3) | 35.1 (26.7) | 32.4 (23.6) | 27.5 (26.0) | 0.127 |
| Longitudinality | 57.1 (25.2) | 59.6 (24.6) | 44.9 (23.9) | 49.5 (27.6) | <0.001 |
| Coordination | 56.2 (30.6) | 57.7 (30.9) | 50.8 (26.2) | 48.4 (32.9) | 0.026 |
| Comprehensiveness (services available) | 51.2 (26.3) | 53.4 (26.2) | 42.3 (24.0) | 41.3 (27.1) | <0.001 |
| Comprehensiveness (services provided) | 36.3 (32.0) | 38.5 (32.3) | 23.5 (27.6) | 33.7 (30.7) | <0.001 |
| Community orientation | 50.2 (19.8) | 51.8 (19.3) | 43.0 (18.8) | 44.0 (23.7) | <0.001 |
| Characteristic | Overall | USC status | P-valuea | ||
|---|---|---|---|---|---|
| (n = 725) | Stable USC (n = 586) | Loss of USC (n = 93) | Change in USC (n = 46) | ||
| Age, mean (SD), y | 59.9 (10.1) | 60.1 (10.0) | 58.0 (10.7) | 61.0 (9.1) | 0.143 |
| Gender, n (%) | 0.029 | ||||
| Female | 383 (52.8) | 303 (51.7) | 47 (50.5) | 33 (71.7) | |
| Educational level, n (%) | 0.068 | ||||
| University or higher | 208 (28.7) | 163 (27.8) | 35 (37.6) | 10 (21.7) | |
| Data missing | 16 (2.2) | 12 (2.0) | 4 (4.3) | 0 (0.0) | |
| Annual household income, n (%), million JPY | 0.421 | ||||
| <3.00 (≒ US$21,000) | 143 (19.7) | 119 (20.3) | 19 (20.4) | 5 (10.9) | |
| 3.00–6.99 | 414 (57.1) | 326 (55.6) | 59 (63.4) | 29 (63.0) | |
| ≥7.00 | 160 (22.1) | 135 (23.0) | 14 (15.1) | 11 (23.9) | |
| Data missing | 8 (1.1) | 6 (1.0) | 1 (1.1) | 1 (2.2) | |
| Number of chronic conditions, mean (SD) | 1.6 (1.4) | 1.6 (1.4) | 1.1 (1.4) | 1.9 (1.4) | 0.003 |
| Data missing, n (%) | 22 (3.0) | 16 (2.7) | 5 (5.4) | 1 (2.2) | |
| EQ-5D-5L, mean (SD) | 0.88 (0.09) | 0.88 (0.10) | 0.89 (0.06) | 0.89 (0.06) | 0.216 |
| Data missing, n (%) | 2 (0.3) | 2 (0.3) | 0 (0.0) | 0 (0.0) | |
| JPCAT | |||||
| Total score, mean (SD) | 47.5 (17.0) | 49.3 (16.9) | 39.5 (13.8) | 40.7 (18.1) | <0.001 |
| Domain scores, mean (SD) | |||||
| First contact | 34.3 (26.3) | 35.1 (26.7) | 32.4 (23.6) | 27.5 (26.0) | 0.127 |
| Longitudinality | 57.1 (25.2) | 59.6 (24.6) | 44.9 (23.9) | 49.5 (27.6) | <0.001 |
| Coordination | 56.2 (30.6) | 57.7 (30.9) | 50.8 (26.2) | 48.4 (32.9) | 0.026 |
| Comprehensiveness (services available) | 51.2 (26.3) | 53.4 (26.2) | 42.3 (24.0) | 41.3 (27.1) | <0.001 |
| Comprehensiveness (services provided) | 36.3 (32.0) | 38.5 (32.3) | 23.5 (27.6) | 33.7 (30.7) | <0.001 |
| Community orientation | 50.2 (19.8) | 51.8 (19.3) | 43.0 (18.8) | 44.0 (23.7) | <0.001 |
aP-value by one-way analysis of variance for continuous variables and chi-square test for categorical data.
Primary care attributes and usual source of care status
Figure 2 shows the results of the multivariable multinominal logistic regression analyses to examine the associations of JPCAT total and domain scores with the loss and change of USC, in reference to stable USC. After adjustment for possible confounders, the JPCAT total score was associated with a decrease in the loss of USC (adjusted relative risk ratio [aRRR] per 10-point increase = 0.70, 95% CI: 0.60–0.81). Regarding the JPCAT domains, longitudinality (aRRR per 10-point increase = 0.80, 95% CI: 0.73–0.88), comprehensiveness (services available) (aRRR per 10-point increase = 0.85, 95% CI: 0.77–0.93), comprehensiveness (services provided) (aRRR per 10-point increase = 0.85, 95% CI: 0.78–0.92), and community orientation (aRRR per 10-point increase = 0.78, 95% CI: 0.69–0.88) were significantly associated with a decrease in the loss of USC.
Multivariable multinomial logistic regression analyses of associations of JPCAT scores with loss and change of USC, in reference to stable USC. Adjusted relative risk ratio per 10-point increase. Adjusted for age, gender, educational level, annual household income, number of chronic conditions, and EQ-5D-5L score.
The JPCAT total score was also associated with a decrease in the change in USC (aRRR per 10-point increase = 0.73, 95% CI: 0.60–0.88). Regarding the JPCAT domains, longitudinality (aRRR per 10-point increase = 0.86, 95% CI: 0.76–0.97), comprehensiveness (services available) (aRRR per 10-point increase = 0.83, 95% CI: 0.73–0.93), and community orientation (aRRR per 10-point increase = 0.80, 95% CI: 0.68–0.94) were significantly associated with a decrease in the change in USC.
Discussion
The present study, using data from a nationwide cohort study, found that high-level primary care attributes were associated with reductions in both loss of USC and change in USC. From a policy perspective, our findings suggest that enhancing the level of primary care attributes may contribute to an increase in the number of residents with stable USC, which potentially leads to decreasing unmet healthcare needs and improving healthcare processes and outcomes. Also, interpreting this study from a clinical practice and management perspective, residents tend to defect from providers with lower levels of primary care attributes, providing a rationale for individual providers and managers to strengthen these functions through quality improvement.
The proportion of participants who changed their USC in this study was smaller than in the USA and larger than in the UK.8,15 On the other hand, the proportion of participants who lost their USC was greater than in the previous study in the United States.14 There are no studies of USC loss in Japan in the past, thus comparisons over time cannot be made, but a possible cause of the high rate of USC loss may be the influence of the COVID-19 pandemic. For example, our prior study showed that approximately one-fifth of adult Japanese residents with a USC reported that their USCs would not accept care for patients with suspected or confirmed COVID-19 infection.21 The accumulation of these pandemic-specific negative experiences may have led to the high rate of USC loss.
Consistent with our findings, a previous cross-sectional study conducted in China reported a positive association between primary care attributes and the presence of USC.13 In addition, another longitudinal study showed that primary care practices with low levels of patient satisfaction, particularly in physician–patient communication and physician trust and confidence, were more likely to have high rates of change in USC.15 Our study provides additional evidence on primary care provider factors associated with the stability of USC status based on a nationwide cohort study.
In the present study, among the attributes of primary care, longitudinality, comprehensiveness (services available), and community orientation were commonly associated with decreased USC loss and USC change. The findings suggest that these attributes are particularly important to individuals’ selection of the USC. In terms of comprehensiveness, it may represent the growing need for primary care to address complex health issues, including psychosocial factors.28 The results for community orientation may reflect the policy emphasis on home medical care in Japan and the expectations placed on primary care providers to engage in public health practice during the pandemic.29
During the COVID-19 pandemic, the Japanese government recognized the need to clarify and strengthen the role of primary care physicians, and in 2023 legislated the function of kakaritsukei, which reflects the attributes of primary care including first contact, longitudinality, coordination, comprehensiveness, and community orientation.30 A new system for primary care institutions to report these functions to prefectural governments is scheduled to be implemented in 2025, and further effective measures are needed to improve the quality of primary care in Japan.
To the best of our knowledge, this is the first longitudinal study to report on the role of primary care attributes in reducing the loss or change of USC. A major advantage of our study is the use of longitudinal data from a nationwide survey that included a sample representative of the adult population of Japan. Another strength of this study was the high follow-up rate. In addition, the JPCAT used in this study is a validated and internationally recognized instrument for assessing primary care attributes.
However, the present study had several potential limitations. First, we did not collect data on the detailed reasons for the loss or change of USC. Therefore, it was not possible to analyze the data for each of these reasons. The inclusion of USC loss due to participant relocation or other involuntary reasons might have underestimated the association between primary care attributes and USC loss. Second, we collected data using a structured questionnaire; however, self-reported data to identify chronic conditions may have introduced a misclassification bias. Third, the sample of participants did not include patients with advanced dementia or other conditions that would prevent them from completing the survey themselves. In addition, there may be differences in the use of primary care resources by residents during and after the COVID-19 pandemic. Therefore, caution should be exercised in generalizing the results of this study.
Conclusion
Our study indicates that primary care attributes play an important role in preventing the loss or change of USC and contribute to the stability of USC status. These findings provide additional rationale for policymakers, healthcare providers, and managers to seek to strengthen core attributes of primary care.
Supplementary material
Supplementary material is available at Family Practice online.
Conflict of Interest
Drs Aoki and Matsushima received lecture fees and lecture travel fees from the Center for Family Medicine Development of Japanese Health and Welfare Co-operative Federation. Drs Aoki and Matsushima are advisers of the Centre for Family Medicine Development practice-based research network. Dr Matsushima’s son-in-law worked at IQVIA Services Japan K.K. which is a contract research organization and a contract sales organization. Dr Matsushima’s son-in-law works at Syneos Health Clinical K.K. which is a contract research organization and a contract sales organization.
Funding
This work was supported by JSPS KAKENHI Grant Number JP20K18849 and Pfizer Health Research Foundation Grant Number 21-E-01.
Ethical approval
The institutional review board of the Jikei University School of Medicine (approval no. 32-416(10505)).
Data Availability
The datasets generated and analyzed in the current study are not publicly available because we did not obtain informed consent from the participants to share the data. However, the datasets are available from the corresponding author upon reasonable request.
