BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.
When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.
The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.
BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).
If you have any questions on BMJ Open’s open peer review process please email [email protected] BMJ Open
Workforce crisis in primary health care worldwide: the Hungarian example in a longitudinal follow-up study
Journal: BMJ Open ManuscriptFor ID peerbmjopen-2018-024957 review only Article Type: Research
Date Submitted by the 03-Jul-2018 Author:
Complete List of Authors: Papp, Magor; National Public Health Institute Kőrösi, László; National Institute of Health Insurance Fund Management Sándor, János; Faculty of Public Health, University of Debrecen, Department of Preventive Medicine Nagy, Csilla; Public Health Administration Service of Government Office of Capital City Budapest Juhász, Attila; Public Health Administration Service of Government Office of Capital City Budapest Ádány, Róza; University of Debrecen, Faculty of Public Health, Department of Preventive Medicine; University of Debrecen, Debrecen, Hungary, MTA-DE Public Health Research Group of the Hungarian Academy of Sciences
PRIMARY CARE, workforce crisis, vacancy, general practitioners, Keywords: socioeconomic status, deprivation
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 19 BMJ Open
1 2 3 4 Workforce crisis in primary health care worldwide: 5 the Hungarian example in a longitudinal follow-up study 6 7 8 Magor Papp1, László Kőrösi2, János Sándor3, Csilla Nagy4, Attila Juhász4, Róza Ádány3,5 9 10 1 National Public Health Institute, Budapest, Hungary 11 12 2 National Institute of Health Insurance Fund Management, Budapest, Hungary 13 14 3 Department of Preventive Medicine, Faculty of Public Health, University of Debrecen, Debrecen, Hungary 15 16 4 Public Health Administration Service of Government Office of Capital City Budapest, Budapest, Hungary 17 18 5 MTA-DE Public HealthFor Research peer Group of the reviewHungarian Academy onlyof Sciences, University of Debrecen, Debrecen, Hungary 19 20 21 Corresponding author: Róza Ádány, MTA-DE Public Health Research Group, 22 23 Department of Preventive Medicine, Faculty of Public Health, University of Debrecen, 24 25 Hungary 26 Debrecen, Kassai str 26, 27 28 H-4028 29 e-mail: [email protected] 30 Phone: +3652512764 31 32 Fax: +3652417267 33 34 35 Word count: 4335 36 37 Keywords: primary care, workforce crisis, vacancy, general practitioners, 38 39 socioeconomic status, deprivation 40 41 42 ABSTRACT 43 44 45 Objective To explore the development of the GP shortage as a secular trend and to 46 examine its characteristics from the perspective of an entire country, Hungary. 47 48 Design Longitudinal follow-up study over the decade 2007-2016. 49 50 51 Methods Analyses were performed on changes in number, age and sex of GPs by year 52 as well as on their geographical distribution and migration between areas categorized 53 into quintiles on the basis of the degree of socioeconomic deprivation. 54 55 56 Setting and subjects The study involved all general practices and GPs with territorial 57 supply obligations for Hungary in the period examined. 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 19
1 2 3 Results The analysis has shown a decrease in the number of GPs in all types of 4 5 practices during the examined decade, especially in the number of GPs in mixed and 6 paediatric practices. The results indicate both ageing and feminization of the GP 7 8 population. The distribution pattern of deprivation correlates well with the relative 9 vacancy rates, as it was confirmed by risk analysis showing exponential association 10 between relative vacancy rate and deprivation. The migration of GPs from the most- 11 12 deprived municipalities to the least-deprived ones was a definitive trend in the period 13 examined. 14 15 16 Conclusions The workforce crisis in primary care is progressively deepening, and 17 departmental intervention is needed. The deepening GPs workforce crisis is more 18 characteristic of theFor most-deprived peer areas review of the country. only The migration of GPs from 19 20 the most-deprived areas to the least-deprived ones further deepens the inequity in 21 access to health care for the most-vulnerable population groups and, in this way, 22 23 usurps their fundamental human right to health. These observations raise the decision- 24 making responsibility to support effective recruitment and retention of GPs for the 25 26 most-deprived areas. 27 28 29 STRENGTHS AND LIMITATIONS 30 31 32 - The study is based on a comprehensive longitudinal follow-up survey at a 33 national level to evaluate major changes in the number and distribution of 34 general practitioners (GPs). 35 - The survey used an area-based composite indicator to study the relationship 36 37 between deprivation and shortage in GPs in the country 38 - The study analysed the migration of GPs between practices, but the reasons for 39 leaving primary care were not investigated in detail. 40 - Data were available for the first day of each year; thus the study could not 41 42 detect transient changes in a year. 43 44 45 INTRODUCTION 46 47 48 The workforce crisis in primary care (PC) is a worldwide phenomenon. The shortage 49 in general practitioners (GPs) is particularly severe in low- and middle-income 50 51 countries in Africa, Asia and the Pacific,1-3 but it is becoming more striking in high- 52 income countries, as well. The Association of American Medical Colleges (AAMC) 53 54 projected a shortage of 46,000 primary care physicians by 2025,4 and it is estimated 55 by Petterson et al in a study designed to calculate the projected primary care physician 56 shortage at the current primary care production rates5 that if it remains unchanged, 57 58 the result will be a shortage in excess of 33,000 primary care physicians by 2035. In 59 the UK, research concluded that NHS England had substantially under-estimated the 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 19 BMJ Open
1 2 3 current shortage of GPs; according to this analysis, in 2016, England was already 4 5 approximately 6,500 GPs below the ideal number, a gap that would increase to 12,100 6 by 2020.6 According to recently reported findings from a regional survey, 82.0% of 7 8 responding GPs stated that they intend to leave general practice, take a career break 9 and/or reduce clinical hours of work within the next 5 years.7-8 Switzerland is also 10 11 facing an impending primary care workforce crisis since almost half of all primary 12 care physicians are expected to retire in the next decade.9-10 As the latest “Health at 13 14 a glance, 2016” report11 shows in the Organisation for Economic Co-operation and 15 Development (OECD)countries ― among them in the EU member states ― the 16 17 workforce crisis at the level of primary care is so deep that in the period of 2011-2013 18 a significant portionFor of thepeer emergency review department only (ED) visits occurred because 19 primary care was not available. In certain Central Eastern European (CEE) countries 20 21 (Czech Republic, Slovakia), this proportion of ED visits was as high as 52-74%. The 22 report concludes that in all countries, especially in CEE countries, there is a need to 23 24 further improve access to and the quality of primary care for the whole population. 25 26 The shortage of GPs seems to be the key issue of the workforce crisis in primary care 27 throughout the world. Data are available almost exclusively from cross-sectional rather 28 29 than longitudinal surveys;6, 12-14 therefore, these studies are unable to report on 30 previous and actual translations in the primary care system. In addition, these surveys 31 32 were conducted among GPs typically in a single region of a country,13, 15-18 which 33 may not be entirely typical of a country as a whole. 34 35 36 Our study on the development of the GP shortage and its characteristics in Hungary 37 illustrates a prime example of CEE countries with a workforce crisis in primary care, 38 and it is designed to characterize the secular trend of the changes in number and 39 40 distribution of GPs by age and sex in a longitudinal follow-up design, as well as to 41 describe the distribution of vacant GP positions by deprivation. 42 43 44 45 METHODS 46 47 Administratively, Hungary is divided into 19 counties as well as the capital Budapest; 48 49 thus, it has 20 European regions at the third level of the Nomenclature of Territorial 50 Units for Statistics (NUTS). The counties are further subdivided into 198 districts 51 52 constituting local administrative units 1 (LAU1), formerly known as NUTS level 4 of 53 Hungary1. The number of local administrative units 2 (LAU2), called municipalities or 54 settlements, in the period examined was 3176. This study focused on the changes in 55 56 the number and distribution of GPs by year, age, sex and deprivation in Hungary in the 57 10-year period that began on 1 January 2007 and ended on 1 January 2017. 58 59 60 Data on general practices analysed
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 19
1 2 3 The analysis is based on data provided by the National Institute of Health Insurance 4 5 Fund Management (NIHIFM) for the period of 1 January 2007 – January 2017 in 6 Hungary. The NIHIFM provided detailed information for the first day of each year on 7 8 the number, type (adult, paediatric or mixed) and geographical (municipal) location of 9 general practices, identified by numeric practice codes. It was reported whether the GP 10 post of the practice was filled or unfilled on the first day of each year. In the case of 11 12 practices with filled GP posts, the age and sex of the GPs were registered. If the GP 13 post was unfilled, the period during which there was an inability to recruit GPs was 14 15 also indicated. If the reason for vacancy involved moving a GP from one primary care 16 practice to another, the deprivation status of the municipalities where the former and 17 18 the newly engagedFor practices peer were foundreview was also only considered in the analysis; the 19 number of practices affected was defined by deprivation index (DI) quintiles. 20 21 In the analysis performed, the change in the number of practices with filled and 22 23 unfilled GP posts and the change in the average age of GPs were defined by the type 24 of practices for the period investigated. In addition, the change in distribution by sex 25 26 was also determined. 27 28 Deprivation 29 30 To see the relationship between the distribution of unfilled practices and the socio- 31 32 economical status (SES) of the settlements where the practices are located, association 33 analyses were carried out. The SES was characterized by the DI, an area-based 34 composite indicator at the municipality level. The DI was built from seven different 35 36 indicators (income, the level of education, the rate of unemployment, the rate of one- 37 parent families, the rate of large families, the density of housing and car ownership) 38 39 derived from the Regional Informational System of the Ministry of Local Government 40 and Regional Development. The data were originally obtained from the Hungarian 41 42 Central Statistical Office (from the Census 2011) and the Hungarian Tax and Financial 43 Control Administration (for the year of the census). The method to calculate DI values 44 45 has been published previously19 and successfully applied in several studies.20-23 46 The development of the DI was carried out by using principal component analysis. All 47 variables were first transformed by using the natural log-transformation and then 48 49 standardised. Areas with positive (higher) index values were municipalities with a 50 lower socio-economic status than the national average, and the converse was shown in 51 52 areas with negative (lower) index values. 53 54 Mapping and risk analysis of vacancy 55 56 A hierarchical Bayesian binomial model was used to predict the relative vacancy ratio 57 58 at the municipality level. The data consisted of the observed number of practices with 59 unfilled GP posts and total number of practices for each municipality. The convolution 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 19 BMJ Open
1 2 3 model, originally proposed by Besag et al,24 was used and incorporates both a 4 5 heterogeneous and spatially structured random effect. The parameters for the spatial 6 model were estimated via Integrated Nested Laplace Approximations (INLA) directly 7 8 in R using the package INLA.25-26 The map for vacancy shows the relative vacancy 9 rate and exceedance posterior probabilities. 10 11 12 Using the risk analysis capabilities of the Rapid Inquiry Facility,27 the association 13 between deprivation and vacancy for GPs was calculated. Relative frequencies of 14 vacancy in relation to the national average were calculated by DI quintiles as a 15 16 municipality-based categorical covariate. Chi-square tests for homogeneity and for 17 linear trend were also carried out to test the global association of the DI and the 18 For peer review only 19 vacancy. 20 21 22 RESULTS 23 24 25 At the beginning of the decade covered by the study, there were 6427 primary care 26 practices operating and serving the country’s population of 10 066 158 (2007); at the 27 28 end of the period, 6350 practices were serving 9 797 561 residents (2016). During the 29 survey period, the number of practices decreased by 77. Behind the change in the 30 number of practices were decisions made by local governments, which are responsible 31 32 for ensuring that primary care services are available to their population. These 33 governments have the right to merge practices that typically could not be fulfilled for 34 35 years in small-sized settlements with other practices belonging to the same 36 administrative municipality. 37 38 Change in number of practices with unfilled GP posts 39 40 41 In Hungary, the number of GPs in the beginning of the decade examined was 6247. Of 42 these, 53.2% served adults, 24.3% served children exclusively, and the remaining 43 44 22.5% served mixed practices. The size of the practices, reflecting the number of 45 clients served, varied widely (800–3000 persons/practice), and the average size was 46 approximately 1550 persons/practice. Generally, more family practices operate in 47 48 more highly populated municipalities, whereas one family practitioner serves more 49 than one municipality in less populated areas. For the end of the period, 6003 GPs 50 51 were registered with the following distribution: 54.6% in adult practices, 23.8% in 52 paediatric practices, and 21.6% in mixed practices (Table 1). 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 19
1 2 3 4 5 2008 2009 2010 6 7 Change 2007-2017 01.01 01.01 01.01 01.01.2007 01.01.2011 01.01.2012 01.01.2013 01.01.2014 01.01.2015 01.01.2016 01.01.2017 8 9 Number of GPs by type of
10 practice 11 Adult practices 3315 3316 3339 3331 3319 3317 3307 3303 3301 3301 3279 -1.1% 12 13 Paediatric practices 1524 1526 1520 1522 1504 1494 1488 1475 1459 1440 1425 -6.5% 14 Mixed practices 1408 1409 1400 1399 1393 1378 1370 1357 1341 1314 1299 -7.7% 15 Total number of GPs 6247 6251 6259 6252 6216 6189 6165 6135 6101 6055 6003 -3.9% 16 Total number of general 17 6427 6415 6417 6408 6395 6382 6383 6372 6370 6356 6350 -1.2% practices 18 For peer review only 19 Practices with a vacant GP 20 post 21 Adult practices 45 44 38 38 46 41 53 58 63 64 86 91.1% 22 Paediatric practices 22 16 20 19 30 33 38 46 60 65 71 222.7% 23 113 104 100 99 103 119 127 133 146 172 190 68.1% 24 Mixed practices Total number of practices 25 180 164 158 156 179 193 218 237 269 301 347 92.8% 26 with a vacant GP post 27 Number of practices with a 28 vacant GP post for more 81 93 83 86 93 108 131 142 160 185 240 196.3% 29 than one year 30 Average age of GPs by 31 type of practice 32 Adult practices 53.3 53.6 53.9 54.1 54.6 55.0 55.2 55.7 56.3 56.6 57.0 3.7 yrs 33 34 Paediatric practices 54.1 54.4 55.0 55.6 56.1 56.7 57.1 57.9 58.5 58.9 59.5 5.4 yrs 35 Mixed practices 52.2 52.6 52.8 53.1 53.6 53.7 54.2 54.7 55.3 55.7 56.4 4.2 yrs 36 Sex distribution among 37 GPs 38 Number of female GPs 3265 3291 3317 3344 3349 3359 3377 3383 3390 3383 3373 3.3% 39 40 Number of male GPs 2982 2960 2942 2908 2867 2830 2788 2752 2711 2672 2630 -11.8% 41 42 Table 1: Number and average age of GPs and the number of practices with a 43 44 vacant GP post by the type of practices, as well as, the sex distribution of GPs in 45 Hungary, 2007-2016 46 47 48 49 The number of GPs in all types of practices decreased during the decade, especially 50 the number of GPs in mixed (by 7.7%) and paediatric (by 6.5%) practices. Despite the 51 52 administrative measures resulting in a reduction of the number of practices, the 53 number of practices with a vacant GP post increased dramatically, especially in the 54 55 second half of the period examined. At the beginning of the decade, the number of 56 practices with unfilled GP posts was 180, while at the end of the period it was 347; 57 58 these shifts occurred across 519 settlements. By the end of the decade, the number of 59 adult practices with unfilled GP posts doubled, while the number of paediatric 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 19 BMJ Open
1 2 3 practices with a vacancy for a paediatrician more than tripled. Although the relative 4 5 change in the case of vacant mixed practices was less (61.8% increase), among the 6 practices with a vacant GP post, this type of practice dominated (adult: 24.8%, 7 8 paediatric: 20.5%, mixed: 54.7%), as could be seen at the beginning of the period 9 (adult: 25.0%, paediatric: 12.2%, mixed: 62.8%). The most serious findings were that 10 69.2% of the practices with a vacancy for GP in 2017 had had the vacancy for more 11 12 than a year, and that number was three times higher than it was in 2007 (240 vs 81). 13 14 Change in age and sex distribution of GPs 15 16 The average age of GPs was high even at the beginning of the decade, especially the 17 18 age of the paediatriciansFor peer (54.1 yrs), review but the average only age of GPs serving adult and 19 mixed practices was also over fifty years (53.3 and 52.2 years, respectively). During 20 21 the decade examined, the average age of GPs was increased by 3.7 years in adults, by 22 5.4 yrs in paediatric and by 4.2 yrs in mixed practices (Table 1). Concerning the 23 24 balance by sex, in 2007, 52.3% of the GPs was female, and this rate increased to 25 56.2% by the end of the decade. The age structure graph showing the distribution of 26 female and male GPs by age in 2007 and ten years later in 2017 (Figure 1) clearly 27 28 indicates both ageing and feminization of the GPs population. 29 30 Association between deprivation and practices with an unfilled GP post 31 32 Mapping the DI values shows that the most-deprived areas were found in the north- 33 34 eastern and south-western parts of Hungary. The least-deprived sections were areas in 35 the north-western part of the country and in the capital city of Budapest and its 36 37 neighbouring areas (Figure 2A). The areas of with the highest relative vacancy rates 38 were found along the north-eastern border of Hungary and in the mid-eastern and 39 40 south-western parts of the country (Figure 2B). Maps for DI and practices with vacant 41 GP posts show remarkable overlap; the resultant pattern of excess vacancy rates 42 showed a correlation with a spatial pattern of deprivation, as is verified by the results 43 44 of association analysis for DI quintiles and practices with vacant GP posts (Figure 2C). 45 The results of the risk analysis showed an exponential association between relative 46 2 2 47 vacancy rate and deprivation (χ homogeneity = 232.18, P <0.0010, χ linearity = 48 168.87, P<0.001). 49 50 Based on DI values, in the areas of highest deprivation (quintile V), nearly a fifth 51 52 (18.31% [15.83 - 21.16]) of the practices had a vacancy for a GP; this rate is nearly 53 triple that of the Hungarian average vacancy rate (6.39% [5.88 - 6.94]). Even in 54 55 quintile IV of the municipalities, 8.21% of the practices had vacant GP posts, i.e., the 56 vacancy rate exceeded the national average by 29%. In the least-deprived areas, the 57 58 vacancy rate was 4.07%, which was only 64% of the national average. The vacancy 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 19
1 2 3 rates between the areas of deprivation quintiles, I, II and III did not differ significantly 4 5 (Table 2). 6 7 Total number of 8 Deprivation quintiles settlements with Vacancy rate (%) Relative vacancy rate 9 vacant GP post, 2017 10 D I (least-deprived 11 51 0.64 [0.47 - 0.84] 12 area) 4.07 [3.05 - 5.32] 13 D II 84 3.95 [3.17 - 4.88] 0.62 [0.49 - 0.77] 14 15 D III 107 4.52 [3.76 - 5.44] 0.71 [0.59 - 0.85] 16 D IV 128 8.21 [6.96 - 9.69] 1.29 [1.08 - 1.53] 17 D V (most-deprived 149 2.86 [2.44 - 3.36] 18 area) For peer review18.31 [15.83 only - 21.16] 19 20 Hungary 519 6.39 [5.88 - 6.94] 1 21 22 Table 2: Distribution of settlements with vacant GP posts and the vacancy rates 23 compared to the national average by DI quintiles on 1 January 2017 24 25 26 27 28 Moving GPs to another primary care practice 29 30 In the period examined, a definitive movement of GPs from the more deprived 31 municipalities to the least-deprived ones could be detected (Table 3). In the least- 32 33 deprived (DI quintile I) areas, the number of GPs increased by 2.2%, while in the 34 most-deprived areas (DI V) it decreased by 8.4%. Municipalities in the DI IV and V 35 36 areas were losing GPs while those in DI I, II and III were gaining them. The resultant 37 outcome of this trend is that the distribution of GPs, as well as the access to primary 38 39 care services by DI quintiles, became more inequitable (Table 3). 40 Total Number of Gain/loss in 41 Number of 42 Deprivation quintiles number of incoming number of Change outgoing GPs 43 GPs, 2017 GPs GPs 44 45 D I 809 53 71 18 2.2% 46 D II 1802 98 118 20 1.1% 47 D III 2029 120 142 22 1.1% 48 D IV 1259 123 101 -22 -1.7% 49 50 D V 451 93 55 -38 -8.4% 51 Total 6350 487 487 52 53 54 Table 3: Migration of general practitioners among areas with different 55 deprivation status (DI-DV) between 1 January 2007 and 1 January 2017 56 57 58 59 60 DISCUSSION
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 19 BMJ Open
1 2 3 Primary care is considered to be the foundation of the health care system throughout 4 5 the world. Currently, in countries with rapidly ageing populations and a high number 6 of people living with chronic diseases, a strong and accessible base of primary care 7 8 providers is essential to the proper level of medical services as well to health 9 promotion for the whole population. There is a general agreement in the scientific 10 literature regarding a workforce crisis at the level of primary care; the shortage of 11 12 primary care physicians now calls for attention and urgency worldwide.28 13 14 The WHO, after publishing “World health report: primary care now more than ever” in 15 16 200829 through its Department of Human Resources for Health, intended to provide 17 an overview of the issues, challenges and potential solutions to the problem of 18 For peer review only 19 inequitable access to primary care. The underserved areas were defined as remote and 20 rural areas using the definition of the United Nations, namely, that remote and rural 21 22 areas are “all the areas what are not urban”. The acceptance of this definition suggests 23 that their disadvantageous position is simply the result of their geographical location. 24 Nevertheless, it is documented in the survey that “the deepest concerns of health 25 26 workers in regard to practising in remote and rural areas are those related to the socio- 27 economic environment, such as working and living conditions, access to education for 28 29 children, availability of employment for spouses, insecurity, and work overload.”30 30 This is in good harmony with previous observations demonstrating that the so-called 31 32 “underdoctored” areas are the most-deprived regions of the UK, i.e., “the general 33 practice workforce is unevenly spread across the country, with the fewest doctors in 34 35 the most-deprived areas, exacerbating health inequalities.”31 It is worth mentioning 36 that the conclusion is based on previous studies that have focused on large areas and 37 have not been able to accurately describe socioeconomic inequality in the primary care 38 39 supply. 40 41 In our present report, the following trends and phenomena are clearly identified as 42 43 characteristics of the Hungarian primary care workforce crisis: 44 45 - Increasing number of primary care practices with a vacant GP post 46 - Increasing average age of GPs in practice 47 48 - Feminization of the profession 49 - Strong association between deprived areas and practices with unfilled posts for 50 51 GP 52 - Intense movement of GPs from the deprived areas to the least-deprived ones 53 54 General practitioners provide a complexity of care (screening, health status 55 56 assessment, diagnosis, treatment, follow-up and referral to the secondary level of care, 57 i.e., gatekeeping) through insurance coverage for this basic health care service. 58 59 Although the usefulness and effectiveness of the gatekeeping function of GPs in 60 reducing utilization of emergency and hospital services is a subject of intense
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 19
1 2 3 discussion with contradictory conclusions,see reviewed in 32-33 in the CEE 4 5 countries with a very unfavourable health status of their population,34-35 it seems to 6 be essential. In these countries (except the Czech Republic and Slovakia), patients do 7 8 not have direct access to hospital consultants, but the GPs control access to secondary 9 care.36 Although, currently, more outpatient specialist services can be approached in 10 11 Hungary without referral, the traditional referral system is still the dominant one in 12 practice. There is a trend towards introducing more group practices in the majority of 13 14 the countries in Europe, but solo practice is still the norm in Hungary, Bulgaria, the 15 Czech Republic, Slovakia and Latvia.37 In addition, Hungary, Slovakia and Poland 16 17 are scored as the countries with the weakest primary care governance in Europe.38) 18 Consequently, theFor increasing peer number reviewof practices with only unfilled GP posts may strongly 19 20 affect the health status of the populations by missing preventive services, as well as 21 through a backlog of referrals to relevant specialists. 22 23 The chronic shortage of doctors in the primary health care sector owing to low pay, 24 25 restricted opportunities for training or career progression and poor working conditions 26 has been a well-known fact for a long time,39-40 but in the last decade, an 27 28 exponential increase in vacant GP posts was observed in different countries. Although 29 the NHS England says their data are not robust as the sample size was too small, the 30 Pulse magazine survey of 860 GPs in May 2017 found that 12.2% of positions were 31 32 currently vacant in England, an increase from 2.1% in 2011 when Pulse started 33 collecting these data.41 Similarly, in a survey in Scotland, 22% of responding 34 35 primary care practices reported current GP vacancies as of 31 August 2015, an 36 increase from 9% of practices reporting current vacancies in 2013.42 The same 37 38 phenomenon can be detected in Hungary, where the number of primary care practices 39 with an unfilled GP post almost tripled in the period of 2010-2017. 40 41 42 Parallel with the decrease in the number of GPs, the average age of the GPs became 43 significantly higher. The age distribution clearly indicates that the number of male GPs 44 under the age of 60 years was lower in 2017 than in 2007 across all ages, while in the 45 46 case of female GPs, although the trend is very similar, increases can be seen at certain 47 ages, mainly at younger ones. This fact is in harmony with the observed feminization 48 49 of the profession. Authors from Scotland ― where the primary care workforce 50 changes are similar to those we observed in Hungary ― suppose that the rapidly 51 52 increasing proportion of women in general practice may lead to an increasing shortfall 53 of medical availability in the future if current work patterns are maintained.43 54 55 Nevertheless, in a systematic review based on 32 relevant publications to examine the 56 evidence that quantifies the effect of feminization on practice characteristics, Hedden 57 and her colleagues44 note “that female GPs self-report fewer hours of work than 58 59 male GPs, have fewer patient encounters, and deliver fewer services, but spend longer 60 with their patients during a contact and address more separate presenting problems in
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 19 BMJ Open
1 2 3 one visit”, and conclude that “the feminization of the workforce is likely to have a 4 5 small negative impact on the availability of primary health care services, and the 6 drivers of observed differences between male and female GPs are complex and 7 8 nuanced.” If Hungary is considered among the 10 countries that have the best parental 9 leave policies in the world ― the length of paid maternity leave is 24 weeks and is 10 paid at 70% of salary, which can be extended by an additional time off paid at 40% of 11 12 salary for 136 weeks45 ― female doctors may stay away from work for a fairly long 13 time, which can make finding a suitable temporary replacement problematic. 14 15 16 Behind the fact that in Hungarian primary care there is a strong association between 17 deprived areas and practices with unfilled posts for GPs and the intense movement of 18 For peer review only 19 GPs from the deprived areas to the least-deprived ones is the probability of an existing 20 financial background. Although the majority of GPs are private entrepreneurs 21 contracted by the National Health Insurance Fund Administration (NHIFA) and they 22 23 must send itemized performance reports to the NHIFA, capitation-based financing is 24 almost exclusively based on the list size of the practices, and only approximately 3% 25 26 of GPs’ reimbursement is pay-for-performance, i.e., the level of reimbursement does 27 not depend on the quantity and/or quality of the provided services. Consequently, GPs 28 29 are financially motivated to maximize the number of persons in the practice and 30 minimize performance.46 As can be calculated using data from the financing system, 31 32 the average amount of reimbursement based on capitation is approximately 5000 33 EUR/month/practice, and it must cover maintaining the infrastructure, paying for 34 energy utilization, cleaning, expenses related to using a car, phone bills, salaries of 35 36 other health staff (at least a nurse), pension fund contributions, social and liability 37 insurance, specialists’ services utilized, taxes related to private business, etc. 38 39 According to the estimated data from the Hungarian Medial Chamber, in 2017, the 40 remaining income that can be used to pay the salary of the GP was approximately 41 42 1000 EUR before taxation. Although it can be – and should be – refused on the basis 43 of ethical considerations, it remains a fact that the informal so called “gratuity” 44 payment from the patients for GP services is more or less essential to make a living. 45 46 According to the results of a recent survey carried out by the Residents and Medical 47 Specialists Union in Hungary, the practice of giving “gratuity” money to medical 48 49 doctors could be eliminated only if the doctors’ wages are increased significantly. 50 More than 53% of respondents completely agreed, and 27% mostly agreed with the 51 52 statement “If basic wage increases reached a level indicated by me, then I would 53 consider it proper for the giving and receiving of ‘gratuity money’ – in all its forms – 54 55 to be qualified as corruption and therefore punishable by law.”47 Furthermore, 56 moving GPs from deprived areas to least-deprived ones is ― at least in part ― 57 motivated by the expected increase in “gratuity” payment. As it is reported by Møller 58 59 Pedersen et al, the annual GP income level is typically above that of senior hospital 60 consultants in Denmark. It is interpreted as a deliberate policy to attract and retain
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 19
1 2 3 GPs. “The reasoning is that, although being a GP may not be as prestigious as being a 4 5 cardiac surgeon, there at least should be an added monetary reward.” It is also worth 6 mentioning that approximately one third of a GP's income comes from capitation 7 8 payment based on the number of clients on his or her lists, and two-thirds comes from 9 fee-for-service payments.48 The OECD report now says that “Danish primary care 10 11 has a solid, highly professional and motivated base, and no major changes to its broad 12 configuration are called for.”49 13 14 15 These recruitment and retention problems are well known in the primary care systems 16 of developed countries, but information about effective strategies and interventions is 17 18 very limited. Recently,For apeer systematic review review was undertakenonly in which 51 studies 19 assessing 42 interventions categorized into 13 groups (financial incentives, recruiting 20 21 rural students, international recruitment, rural- or primary-care-focused undergraduate 22 placements, rural or underserved postgraduate training, well-being or peer support 23 24 initiatives, marketing, mixed interventions, support for professional development or 25 research, retainer schemes, re-entry schemes, specialized recruiters or case managers 26 and delayed partnerships) were retrieved. Weak evidence supported the use of 27 28 postgraduate placements in underserved areas, undergraduate rural placements and 29 recruiting students to medical school from rural areas, and there was mixed evidence 30 31 about financial incentives.50 32 33 A single report on a study using lower super output area data and the Index of Multiple 34 Deprivation 2010 rankings for England described substantial inequalities in GPs’ 35 36 service, expressed in full-time equivalent (FTE) GPs before the ‘Equitable Access to 37 Primary Medical Care’ (EAPMC) programme was launched in December 2007. 38 39 EMPMC invested £250 million towards establishing new general practices and GP-led 40 heath centres, as well as towards extending opening hours and expanding services in 41 42 the 38 most “underdoctored” Primary Care Trust areas and providing a guaranteed 43 competitive salaries for GPs. Total numbers of FTE for GPs have grown much faster 44 45 in the most-deprived fifth of small areas in England than elsewhere, with the GP 46 supply in the most affluent fifth growing at the slowest pace over the 10 years covered 47 by the study.51 48 49 50 The limitation of our study is that it is an ecological one that assumes that individual 51 members of a group all have the average characteristics of the group as whole, when, 52 53 in fact, any association observed between variables at the group level does not 54 necessarily mean that the same association exists for an individual plucked from the 55 group. In addition, we had no opportunity to study the effects of different health care 56 57 reforms on the primary care workforce in Hungary. Although efforts have been made 58 to reduce hospital activities and to improve primary care services, health care 59 60 provision is still highly hospital-centred. While the number of acute care hospital beds
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 19 BMJ Open
1 2 3 has decreased by 30% since 2000, the total number of hospital beds remains above the 4 5 EU average (699 compared to 515 beds per 100 000 population in 2015). In 2009, a 6 pay-for-performance system was introduced to incentivise ‘good’ primary care. More 7 8 recent legislation in 2015 also tried to strengthen primary care by redefining GPs’ 9 tasks (including an emphasis on disease prevention), promoting community practices 10 11 and health promoting offices, and revising payment schemes.52 Despite these 12 reforms, primary care is still mainly financed by capitation and a fixed budget, and no 13 permanent incentives were introduced to support movement by GPs into underserved 14 15 deprived areas. However, since 2014, a single settlement allowance has been given to 16 GPs at the time of entering into a permanently vacant GP posts. Although studies on 17 18 the effect of financialFor incentives peer for reviewGPs in Hungary only are lacking, it is reasonable to 19 suppose that without increasing the salaries of GPs there is no chance to stop the 20 21 deepening workforce crisis in Hungarian primary care. According to OECD figures, 22 the income of a GP in Denmark is five times larger than the income of a GP in 23 Hungary, and the average GP in Scandinavia earns 5-10 times more than the average 24 25 GP in Hungary or in other CEE and Baltic53 countries. With such a large difference, 26 the move not only to other fields of medical specialisation but even to other countries 27 28 in borderline Europe appears easy to justify. In 2013, a salary increase programme for 29 GPs started in Hungary, but the annual growth rate in nominal terms was as low as 30 31 2.3%.54 In addition, pilot projects should be launched immediately to test what kind 32 of interventions targeting the recruitment of GPs and retention can be successfully 33 34 applied in the country to stop the deteriorating trends in primary care. 35 36 37 FIGURE LEGENDS 38 39 40 Figure 1: The age and sex distribution of GPs in 2007 and 2017 in Hungary 41 42 Figure 2: The distribution pattern of deprivation (A) and of the relative vacancy rates 43 (B) in Hungary and the relationship between deprivation and the relative vacancy rate 44 (C) 45 46 47 48 49 REFERENCES 50 51 1. Campbell J, Dussault G, Buchan J, et al. A universal truth: no health without a 52 workforce. Forum report, third global forum on human resources for health, 53 Recife, Brazil. Geneva: World Health Organization, 2013. 54 55 http://www.who.int/workforcealliance/knowledge/resources/ hrhreport2013/en/ 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 19
1 2 3 2. Hays R, Pong LT, Leopando Z, Primary care in the Asia-Pacific region: 4 5 challenges and solutions. Asia Pac Fam Med 2012;11: 8. doi:10.1186/1447- 6 056X-11-8 7 3. Strasser R, Kam SM, Regalado SM, Rural health care access and policy in 8 developing countries. Annu Rev Public Health. 2016;37:395-412. doi: 9 10 10.1146/annurev-publhealth-032315-021507. Epub 2016 Jan 6. 11 4. Berwick DM, Hackbarth AD, Eliminating waste in US health care. JAMA. 12 2012;14:1513–6. 13 5. Petterson, SM, Liaw, WR, Tran, C, et al. Estimating the Residency Expansion 14 15 Required to Avoid Projected Primary Care Physician Shortages by 2035, Ann 16 Fam Med 2015;2: 107–14. doi:10.1370/afm.1760 17 6. Dale J, Potter R, Owen K, et al. Retaining the general practitioner workforce in 18 England: whatFor matters peer to GPs? A review cross-sectional onlystudy. BMC Fam Pract 19 2015;1:140.doi:10.1186/s12875-015-0363-1. 20 21 7. Hayhoe B, Majeed A, Hamlyn M, et al. Primary care workforce crisis: how many 22 more GPs do we need? RCGP Annual Conference, Harrogate, 2016. 23 8. Majeed A, Shortage of general practitioners in the NHS. 2017; BMJ 2017; 24 358:j3191 25 26 9. Burla L, Widmer M, Le corps médical en Suisse: effectif et évolutions jusqu’en 27 2011. Neuchâtel: Observatoire suisse de la santé; 2012. 28 10. Seematter-Bagnoud L, Junod J, Jaccard Ruedin H, et al. Offre et recours aux 29 soins médicaux ambulatoires en Suisse – Projections à l’horizon 2010. Neuchâtel: 30 31 Observatoire suisse de la santé; 2008. 32 11. OECD/EU (2016), Health at a Glance: Europe 2016: State of Health in the EU 33 Cycle, OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264265592-en 34 https://ec.europa.eu/health/sites/health/files/state/docs/health_glance_2016_rep_e 35 36 n.pdf 37 12. Watson J, Humphrey A, Peters Klimm F, et al. Motivation and satisfaction in GP 38 training: a UK cross-sectional survey. Br J Gen Pract 2011; 591: e645–9. 39 13.Fletcher E, Abel GA, 40 http://bmjopen.bmj.com/content/7/4/e015853 - aff-1 41 Anderson R, et al. Quitting patient care and career break intentions among general 42 practitioners in South West England: findings of a census survey of general 43 practitioners. BMJ Open 2017, http://dx.doi.org/10.1136/bmjopen-2017-015853 44 14.Blane DN, McLean G, Watt G, Distribution of GPs in Scotland by age, gender 45 46 and deprivation. Scott Med J 2015;4:214-9. doi:10.1177/0036933015606592. 47 Epub 2015 Sep 23. 48 15. Iacobucci G, GPs in Northern Ireland face "full blown crisis," BMA warns. BMJ. 49 2016 Jun 7;353:i3202. doi: 10.1136/bmj.i3202. 50 51 16. Zieler P, Physician shortage in Saxony-Anhalt. The former youngster returns as 52 country doctor. MMW Fortschr Med. 2015;17:36-7. 53 17. Schmidt S, Gresser U, Development and consequences of physician shortages in 54 Bavaria. Versicherungsmedizin. 2014;1:25-9. 55 56 18. Roeger LS, Reed RL, Smith BP, Equity of access in the spatial distribution of 57 GPs within an Australian metropolitan city. Aust J Prim Health. 2010;4:284-90. 58 doi: 10.1071/PY10021. 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 19 BMJ Open
1 2 3 19. Juhasz A, Nagy C, Paldy A, Development of a deprivation index and its relation 4 5 to premature mortality due to diseases of the circulatory system in Hungary, 6 1998-2004. Soc Sci Med 2010;70:1342–9. 7 20. Nagy C, Juhasz A, Beale L, et al. Mortality amenable to health care and its 8 relation to socio-economic status in Hungary, 2004-08. Eur J Public Health 9 10 2012;22:620–4. 11 21. Nagy C, Juhasz A, Papp Z, et al. Hierarchical spatio-temporal mapping of 12 premature mortality due to alcoholic liver disease in Hungary, 2005-2010. Eur J 13 Public Health 2014;24:827–833. 14 15 22. Boruzs K, Juhasz A, Nagy C, et al. Relationship between statin utilization and 16 socioeconomic deprivation in Hungary. Front Pharmacol 2016;7:66. 17 23. Jakab Zs, Juhasz A, Nagy C, et al. Trends and territorial inequalities of incidence 18 and survival ofFor childhood peer leukaemia review and their relations only to socioeconomic status in 19 Hungary, 1971–2015. Eur J Cancer Prev 1. 20 21 2017;10.1097/CEJ.0000000000000386. 22 24. Besag J, York J, and Mollié A, A Bayesian image restoration with two 23 applications in spatial statistics. Ann Inst Stat Math.1991;43:1–20. doi: 24 10.1007/BF00116466 25 26 25. R Development Core Team. R: A Language and Environment for Statistical 27 Computing, R Foundation for Statistical Computing, Vienna, Austria, 2017; 28 www.R-project.org 29 26. Rue H, S. Martino S, and Chopin N, Approximate Bayesian inference for latent 30 31 Gaussian models using integrated nested Laplace approximations (with 32 discussion). Journal of the Royal Statistical Society, Series B, 2009;2:319-392, 33 www.r-inla.org 34 27. Beale L, Hodgson S, Abellan JJ, et al. Evaluation of spatial relationships between 35 36 health and the environment: the rapid inquiry facility. Environ. Health Perspect 37 2010;118, 1306–12. doi: 10.1289/ehp.0901849 38 28. Song Z, Chopra V, McMahon LF, Addressing the Primary Care Workforce 39 Crisis. Am J Manag Care 2015;8:e452-4. 40 41 29. World Health Organization, The World Health Report - Primary health care (now 42 more than ever). World Health Organization, 2008 43 30. World Health Organization, Increasing access to health workers in remote and 44 rural areas through improved retention. Background paper. Geneva, 2009 45 46 31. Baker M, Ware J, Morgan K, Time to put patients first by investing in general 47 practice. Br J Gen Pract 2014;64:268–9. doi:10.3399/bjgp14X679921 48 32. Van den Heede K, Van de Voorde C, Interventions to reduce emergency 49 department utilisation: A review of reviews. Health Policy 2016;120:1337-1349. 50 51 33. Greenfield G, Foley K, Majeed A, Rethinking primary care's gatekeeper role. 52 BMJ 2016;i4803. doi:10.1136/bmj.i4803 53 34. Karanikolos M, Adany R, McKee M, The epidemiological transition in Eastern 54 and Western Europe: a historic natural experiment. Eur J Public Health 55 56 2017;suppl_4:4-8. doi: 10.1093/eurpub/ckx158 57 35. Mackenbach JP, Nordic paradox, Southern miracle, Eastern disaster: persistence 58 of inequalities in mortality in Europe. Eur J Public Health 2017;suppl_4:14-17. 59 doi: 10.1093/eurpub/ckx160 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 19
1 2 3 36. Euro Health Consumer Index 2016, Health Consumer Powerhouse Ltd, 2017 4 5 37. Sheiman I, Shevski V, Two models of primary health care development: Russia 6 vs. Central and Eastern European countries. National Research University, Higher 7 School of Economics, Moscow, Russia, Series Public and Social Policy, WP BRP 8 06/PSP, 2017 https://wp.hse.ru/data/2017/03/20/1170026603/06PSP2017.pdf 9 10 38. Lember M, Cartier T, Bourgueil Y, Structure and organization of primary care. 11 In: Kringos DS, Boerma WGW, Hutchinson A (eds), Building primary care in a 12 changing Europe, Observatory Studies Series, No. 38, 2015, p. 41-9. 13 39. Rechel B, Dubois C-A, McKee M (eds), The Health Care Workforce in Europe 14 15 Learning from experience. European Observatory on Health Systems and 16 Policies, 2006, 17 40. Kuehn BM, Reports Warn of Primary Care Shortages. JAMA. 2008;300:1872–5. 18 doi: 10.1001/jama.300.16.1872.For peer review only 19 41. NHSProvider: Summary of vacancy, shortfall and fill rate data for the clinical 20 21 workforce. November 2017, https://nhsproviders.org/media/3903/summary-of- 22 vacancy-shortfall-and-fill-rate-data-for-the-clinical-workforce.pdf 23 42. Primary Care Workforce Survey Scotland 2015, A Survey of Scottish General 24 Practices and General Practice Out of Hours Services. National Services Scotland, 25 26 Information Services Division, Publication Report, June 14, 2016 27 http://www.isdscotland.org/Health-Topics/General-Practice/Publications/2016- 28 06-14/2016-06-14-PrimaryCareWorkforceSurveyScotland2015-Report.pdf 29 43. McKinstry B, Colthart I, Elliott K, et al. The feminization of the medical work 30 31 force, implications for Scottish primary care: a survey of Scottish general 32 practitioners. BMC Health Serv Res 2006; 6: 56.doi: 10.1186/1472-6963-6-56 33 44. Heeden L, Barer ML, Cardiff K, et al. The implications of the feminization of the 34 primary care physician workforce on service supply: a systematic review. Human 35 36 Resources for Health 2014;12:32 https://doi.org/10.1186/1478-4491-12-32 37 45. Weller C,These 10 countries have the best parental leave policies in the world, 38 World Economic Forum, 2016 39 https://www.weforum.org/agenda/2016/08/these-10-countries-have-the-best- 40 41 parental-leave-policies-in-the-world, 42 46. Sándor J, Kósa K, Papp M, et al. Capitation-Based Financing Hampers the 43 Provision of Preventive Services in Primary Health Care. Front Public Health 44 2016;4:200, doi: 10.3389/fpubh.2016.00200. eCollection 2016 45 46 47. Spike J: “Gratuity money” for doctors could be eliminated within 4 years, 47 according to doctors union president. Budapest Beacon, March 04, 2018 48 https://budapestbeacon.com/gratuity-money-doctors-eliminated-within-4-years- 49 according-doctors-union-president/ 50 51 48. Møller Pedersen K, Andersen JS, Søndergaard J, General Practice and Primary 52 Health Care in Denmark. J Am Board Fam Med March-April 2012 vol. 25 no. 53 Suppl 1 S34-8 54 49. Forde I, Nader C, Socha-Dietrich K, et al. Primary care review of Denmark, 55 56 OECD Directorate for Employment, Labour and Social Affairs, 2016 57 50. Verma P, Ford JA, Stuart A, et al. A systematic review of strategies to recruit and 58 retain primary care doctors. BMC Health Serv Res 2016;16:126. 59 https://doi.org/10.1186/s12913-016-1370-1 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 19 BMJ Open
1 2 3 51. Asaria M, Cookson R, Fleetcroft R, et al. Unequal socioeconomic distribution of 4 5 the primary care workforce: whole-population small area longitudinal study. BMJ 6 Open 2016;1:e008783. doi: 10.1136/bmjopen-2015-008783 7 52. European Comission, State of Health in the EU, Hungary, Country Health Profile, 8 2017 9 10 53. OECD Health Statistics 2017 11 54. OECD Health Statistics 2015, http://dx.doi.org/10.1787/health-data-en. 12 13 14 FOOTNOTES 15 16 17 Contributors MP, JS, CN, AJ and RA developed the study design. LK identified the 18 practices by location,For vacancy peer and characteristics review of GPs. only MP, AJ and CN did the data 19 processing and analyses. MP wrote the first draft of the paper. All authors critically 20 revised the first draft and approved the final manuscript submitted for publication. 21 22 Funding This work was supported by the Swiss Contribution SH/8/1 project and by 23 the GINOP-2.3.2-15-2016-00005 project financed by the European Union under the 24 European Social Fund and European Regional Development Fund, as well as by the 25 26 Hungarian Academy of Sciences (TK2016-78). 27 28 Competing interests None declared. 29 30 Patient consent Not required. Patients and or public were not involved. 31 32 Ethics approval Not required. 33 34 Provenance and peer review Not commissioned; externally peer reviewed. 35 36 Data sharing statement Data were provided by the National Institute of Health 37 Insurance Fund Management (NIHIFM), the Regional Informational System of the 38 Ministry of Local Government and Regional Development, the Hungarian Central 39 Statistical Office and the Hungarian Tax and Financial Control Administration. Data 40 41 requests should be sent to Professor Roza Adany (corresponding author, 42 [email protected]). 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 19
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 170x175mm (300 x 300 DPI) 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 19 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 170x250mm (300 x 300 DPI) 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open
Workforce crisis in primary health care worldwide: the Hungarian example in a longitudinal follow-up study
Journal: BMJ Open ManuscriptFor ID peerbmjopen-2018-024957.R1 review only Article Type: Research
Date Submitted by the 10-Mar-2019 Author:
Complete List of Authors: Papp, Magor; National Public Health Institute Kőrösi, László; National Institute of Health Insurance Fund Management Sándor, János; Faculty of Public Health, University of Debrecen, Department of Preventive Medicine Nagy, Csilla; Public Health Administration Service of Government Office of Capital City Budapest Juhász, Attila; Public Health Administration Service of Government Office of Capital City Budapest Ádány, Róza; University of Debrecen, Faculty of Public Health, Department of Preventive Medicine; University of Debrecen, Debrecen, Hungary, MTA-DE Public Health Research Group of the Hungarian Academy of Sciences
Primary Subject General practice / Family practice Heading:
Secondary Subject Heading: Public health, Health services research
PRIMARY CARE, workforce crisis, vacancy, general practitioners, Keywords: socioeconomic status, deprivation
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 23 BMJ Open
1 2 3 4 Workforce crisis in primary health care worldwide: 5 the Hungarian example in a longitudinal follow-up study 6 7 8 Magor Papp1, László Kőrösi2, János Sándor3, Csilla Nagy4, Attila Juhász4, Róza Ádány3,5 9 10 1 National Public Health Institute, Budapest, Hungary 11 12 2 National Institute of Health Insurance Fund Management, Budapest, Hungary 13 14 3 Department of Preventive Medicine, Faculty of Public Health, University of Debrecen, Debrecen, Hungary 15 16 4 Public Health Administration Service of Government Office of Capital City Budapest, Budapest, Hungary 17 18 5 MTA-DE Public HealthFor Research peer Group of the reviewHungarian Academy onlyof Sciences, University of Debrecen, Debrecen, Hungary 19 20 21 Corresponding author: Róza Ádány, MTA-DE Public Health Research Group, 22 23 Department of Preventive Medicine, Faculty of Public Health, University of Debrecen, 24 25 Hungary 26 Debrecen, Kassai str 26, 27 28 H-4028 29 e-mail: [email protected] 30 Phone: +3652512764 31 32 Fax: +3652417267 33 34 35 Word count: 4513 36 37 Keywords: primary care, workforce crisis, vacancy, general practitioners, 38 39 socioeconomic status, deprivation 40 41 42 ABSTRACT 43 44 45 Objective The study was designed to explore the development of the general 46 practitioner (GP) shortage in primary care and its characteristics in Hungary. 47 48 Design Longitudinal follow-up study over the decade 2007-2016. 49 50 51 Methods Analyses were performed on changes in number, age and sex of GPs by 52 practice type (adult, paediatric and mixed), as well as on their geographical 53 distribution and migration between areas characterized by deprivation index (DI) at 54 55 municipality level. The association between deprivation and vacancy for GPs was 56 studied by risk analysis. The number of population underserved was defined by DI 57 quintile. 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 23
1 2 3 Setting and subjects The study involved all general practices and GPs in the period 4 5 examined. 6 7 Main outcome measure It is showed that the number of general practices with 8 unfilled GP posts was increasing exponentially, mainly in the most-deprived areas of 9 10 the country. 11 12 Results A decrease in the number of GPs in all types of practices, especially in mixed 13 14 (by 7.7%; p<0.001) and paediatric (by 6.5%; p<0.001) ones was shown; the number of 15 adult practices with unfilled GP posts doubled, while the number of paediatric 16 practices with a vacancy for a paediatrician more than tripled. The average age of GPs 17 18 was increased by For3.7 years peer (p<0.001) review in adult, by 5.4 only yrs (p<0.001) in paediatric and 19 by 4.2 yrs (p<0.001) in mixed practices. In 2007, 52.27% (95%CI 51.03-53.5) of the 20 21 GPs was female, and this rate increased to 56.19% (95%CI 54.93-57.44) by the end of 22 the decade. An exponential association between relative vacancy rate and deprivation 23 24 was confirmed. As a result of the migration of GPs, in the most-deprived areas the 25 number of GPs decreased by 8.43% (95%CI 5.86-10.99). 26 27 Conclusions The workforce crisis in Hungarian primary care is progressively 28 29 deepening and resulting in more severe inequity in access to health care. 30 31 32 STRENGTHS AND LIMITATIONS 33 34 35 - The study is based on a comprehensive longitudinal follow-up survey at a 36 national level to evaluate major changes in the number and distribution of 37 general practitioners (GPs). 38 - The survey used an area-based composite indicator to study the relationship 39 40 between deprivation and shortage in GPs in the country. 41 - The study analysed the migration of GPs between practices, but the reasons for 42 leaving primary care were not investigated in detail. 43 - Data were available for the first day of each year; thus the study could not 44 45 detect transient changes in a year. 46 47 48 INTRODUCTION 49 50 51 The workforce crisis in primary care (PC) is a worldwide phenomenon. The shortage 52 in general practitioners (GPs) is particularly severe in low- and middle-income 53 54 countries in Africa, Asia and the Pacific,1-3 but it is becoming more striking in high- 55 income countries as in the USA,4-5 UK,6-8 Switzerland,9-10 as well. In 2016, 56 57 England was already approximately 6,500 GPs below the ideal number, and this gap 58 would increase to 12,100 by 2020.6 Switzerland is also facing an impending primary 59 60 care workforce crisis since almost half of all primary care physicians are expected to
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 23 BMJ Open
1 2 3 retire in the next decade.9-10 As the “Health at a glance, 2016” report11 shows in 4 5 the Organisation for Economic Co-operation and Development (OECD) countries ― 6 among them in the EU member states ― the workforce crisis at the level of primary 7 8 care is so deep that in the period of 2011-2013 a significant portion of the emergency 9 department (ED) visits occurred because primary care was not available. In certain 10 11 Central Eastern European (CEE) countries (Czech Republic, Slovakia), this proportion 12 of ED visits was as high as 52-74%. 13 14 The shortage of GPs seems to be the key issue of the workforce crisis in primary care 15 16 throughout the world which may strongly contribute to the limited access of people 17 living in socioeconomically deprived conditions to primary care and other health 18 For peer review only 19 services.12 The Position Paper of the Irish Medical Organisation on health 20 inequalities denotes that vulnerable rural and deprived urban communities have no 21 22 adequate GP cover in Ireland.13 Demographic studies identified a shortage of 23 primary care service providers located in the countryside and a relative oversupply in 24 metropolitan areas in Germany, and in German cities with a population size of more 25 26 than 500,000 the distribution patterns of GPs were significantly correlated with the 27 geodemographic index, an area measure of deprivation.14 Although the Royal 28 29 College of General Practitioners reported that the so-called “underdoctored” areas are 30 the most-deprived regions of the UK, i.e., “the general practice workforce is unevenly 31 32 spread across the country, with the fewest doctors in the most-deprived areas, 33 exacerbating health inequalities”,15 it is worth mentioning that there was a 34 35 substantial reduction in socioeconomic inequality in family physician supply 36 associated with national policy for establishing new practices in the most deprived 37 38 areas under the ‘Equitable Access to Primary Medical Care’ programme.[16] 39 40 Substantial geographical inequalities in family physician supply was demonstrated 41 between large subnational areas even in high-income countries with universal health 42 43 coverage. However, because these studies have focused on large areas they have not 44 been able to accurately describe socioeconomic inequality in primary care supply by 45 46 pinpointing primary care shortages in specific disadvantaged neighbourhoods.[17-26] 47 48 The WHO Department of Human Resources for Health provided an overview on the 49 inequitable access of populations to primary care. The underserved populations were 50 51 identified in remote and rural areas where the socio-economic environment, such as 52 working and living conditions, as well as access to education for children were 53 54 unfavourable.27-28 Data are available almost exclusively from cross-sectional rather 55 than longitudinal surveys;6, 29-31 therefore, these studies are unable to report on 56 57 previous and actual translations in the primary care system. In addition, these surveys 58 were conducted among GPs typically in a single region of a country,30, 32-35 which 59 60 may not be entirely typical of a country as a whole.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 23
1 2 3 Concerning the population size of Hungary, as well as the fact that the workforce crisis 4 5 in primary care was evolving in the last decade the country may serve as an ideal field 6 to a study targeting the characterization of the workforce crisis in its development. In 7 8 Hungary responsibilities for primary care have been centralized at national level and 9 the only source of financing is the National Health Insurance Fund. The local 10 governments are responsible for ensuring that primary care services are available to 11 12 their population, so practices are organized by them at municipality level. The size of 13 the practices, reflecting the number of clients served, varies widely (800–3000 14 15 persons/practice), and the average size is approximately 1550 persons/practice. 16 Generally, more family practices operate in more highly populated municipalities, 17 18 whereas one familyFor practitioner peer serves review more than one only municipality in less populated 19 areas. Primary care physicians are mostly (95%) self-employed service providers 20 contracted to the National Health Insurance Fund, and they work almost exclusively in 21 22 solo practices, i.e. one general practitioner - with the assistance of one practice nurse - 23 serves the clients. GPs provide a complexity of care (screening, health status 24 25 assessment, diagnosis, treatment, follow-up and referral to the secondary level of care) 26 through capitation based insurance coverage for this basic health care service. 27 28 Concerning the type of practices, about half of them serves only adults, one quarter of 29 them serves exclusively children, while the remainings are mixed practices. If a GP 30
31 post becomes vacant it is very difficult to recruit replacement GP, in general GPs from 32 the nearby practices as locum-tenents provide services in very limited working hours, 33 so the population belonging to the practice with vacant GP post will be definitively 34 35 underserved. Theoretically for GPs there are some competitive work opportunities, 36 because about 40% of them in addition to specialization in family care have medical 37 38 specialization in paediatrics or internal medicine, but the health care institutions at 39 secondary or tertiary levels rarely provide GPs with positions, so they can work almost 40 41 exclusively in the emergency service at primary care level. On the basis of the results 42 obtained in the Hungarian Public Health Focused Primary Care Model Programme[36- 43 39] the government started to disseminate the GPs’ cluster work in a new pilot 44 45 programme in 2018. GPs’ clusters are built on the cooperation of six GPs employing 46 health professionals (public health professionals, community nurses, physiotherapists, 47 48 dieticians, health psychologists) who – in addition to the traditional patient care - are 49 competent to plan and implement various public health services at cluster level. In 50 51 2019, approximately 6% of the primary care practices is involved in this pilot, i.e. solo 52 practice is still the norm in Hungary. On the basis of the three dimensions of primary 53 care structure – governance, economic conditions and workforce development – the 54 55 Hungarian primary care system is qualified into the medium category,[40] but the 56 coordination and comprehensiveness are found to be weak.[18] 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 23 BMJ Open
1 2 3 Our study is designed to follow the development of the GP shortage and to describe 4 5 the characteristics of changes in Hungary, a country which illustrates a prime example 6 of CEE countries with a workforce crisis in primary care. The secular trend by changes 7 8 in the number, age and sex of GPs by type of practices, as well as the distribution of 9 GPs and vacant GP positions by deprivation are characterized in a longitudinal follow- 10 up design in the 10-year period that began on 1 January 2007 and ended on 1 January 11 12 2017. 13 14 15 16 METHODS 17 18 Data on general practicesFor peer analysed review only 19 20 The analysis is based on data provided by the National Institute of Health Insurance 21 22 Fund Management (NIHIFM) for the period of 1 January 2007 – January 2017 in 23 Hungary. The NIHIFM provided detailed information for the first day of each year on 24 the number, type (adult, paediatric or mixed) and geographical (municipal) location of 25 26 general practices, identified by numeric practice codes. It was reported whether the GP 27 post of the practice was filled or unfilled on the first day of each year. In the case of 28 29 practices with filled GP posts, the age and sex of the GPs were registered. If the GP 30 post was unfilled, the period during which there was an inability to recruit GP, as well 31 32 as the number of population underserved (i.e. the number of people belonging to 33 practice with vacant GP post) were also indicated. If the reason for vacancy involved 34 moving a GP from one primary care practice to another, the deprivation status of the 35 36 municipalities where the former and the newly engaged practices were found was also 37 considered in the analysis; the number of practices affected was defined by deprivation 38 39 index (DI) quintiles. 40 41 In the analysis performed, the change in the number of practices with filled and 42 unfilled GP posts and the change in the average age of GPs were defined by the type 43 44 of practices for the period investigated. In addition, the change in distribution by sex 45 was also determined. 46 47 48 Deprivation 49 50 To see the relationship between the distribution of unfilled practices and the socio- 51 economical status (SES) of the settlements where the practices are located, association 52 53 analyses were carried out. The SES was characterized by the DI, an area-based 54 composite indicator at the municipality level. The DI was built from seven different 55 56 indicators (income, the level of education, the rate of unemployment, the rate of one- 57 parent families, the rate of large families, the density of housing and car ownership) 58 derived from the Regional Informational System of the Ministry of Local Government 59 60 and Regional Development. The data were originally obtained from the Hungarian
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 23
1 2 3 Central Statistical Office (from the Census 2011) and the Hungarian Tax and Financial 4 5 Control Administration (for the year of the census). The development of the DI was 6 carried out by using principal component analysis. All variables were first transformed 7 8 by using the natural log-transformation and then standardised. Areas with positive 9 (higher) index values were municipalities with a lower socio-economic status than the 10 national average, and the converse was shown in areas with negative (lower) index 11 12 values. This method to calculate DI values has been published previously by us41 13 and successfully used in several studies.42-45 Same or similar approaches measuring 14 15 SES were also applied in studies carried out in other countries all over Europe[46] 16 such as the United Kingdom,[47-50] France,[51] Spain,[52-53] Germany,[54] 17 18 Denmark,[55] Switzerland,[56]For peer Italy,[57] review Czech Republic,[58] only and Slovenia.[59] 19 20 21 22 Mapping and risk analysis of vacancy 23 24 25 A hierarchical Bayesian binomial model was used to predict the relative vacancy ratio 26 at the municipality level. The data consisted of the observed number of practices with 27 unfilled GP posts and total number of practices for each municipality. The convolution 28 29 model, originally proposed by Besag et al,60 was used and incorporates both a 30 heterogeneous and spatially structured random effect. The parameters for the spatial 31 32 model were estimated via Integrated Nested Laplace Approximations (INLA) directly 33 in R using the package INLA.61-62 The map for vacancy shows the relative vacancy 34 35 rate and exceedance posterior probabilities. 36 37 Using the risk analysis capabilities of the Rapid Inquiry Facility,63 the association 38 39 between deprivation and vacancy for GPs was calculated. Relative frequencies of 40 vacancy in relation to the national average were calculated by DI quintiles as a 41 municipality-based categorical covariate. Chi-square tests for homogeneity and for 42 43 linear trend were also carried out to test the global association of the DI and the 44 vacancy. 45 46 47 48 RESULTS 49 50 The number of local administrative units 2 (LAU2), called municipalities or 51 52 settlements in Hungary was 3176 with a total population number decreasing from 53 10,066,158 (2007) to 9,797,561 (2017) in the period examined. 54 55 At the beginning of the decade covered by the study, there were 6427 primary care 56 57 practices serving the country’s population, and at the end of the period, 6350 practices 58 were operating. During the survey period, the number of practices decreased by 77. 59 60 Behind the change in the number of practices were decisions made by local
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 23 BMJ Open
1 2 3 governments having the right to merge practices that typically could not be fulfilled 4 5 for years in small-sized settlements with other practices belonging to the same 6 administrative municipality. 7 8 9 10 11 Change in number of practices with unfilled GP posts 12 13 In Hungary, the number of GPs in the beginning of the decade examined was 6247. Of 14 these, 53.2% served adults, 24.3% served children exclusively, and the remaining 15 16 22.5% served mixed practices. For the end of the period, 6003 GPs were registered 17 with the following distribution: 54.6% in adult practices, 23.8% in paediatric practices, 18 For peer review only 19 and 21.6% in mixed practices (Table 1). 20 21 22 2007 2017 23 2008 2009 2010 24
25 01.01.2007 01.01 01.01 01.01 01.01.2011 01.01.2012 01.01.2013 01.01.2014 01.01.2015 01.01.2016 01.01.2017 26 Number of GPs by Change
27 type of practice (p for trend) 28 -1.1% 29 Adult practices
30 3315 3316 3339 3331 3319 3317 3307 3303 3301 3301 3279 (p=0.004) 31 -6.5% 32 Paediatric practices
33 1524 1526 1520 1522 1504 1494 1488 1475 1459 1440 1425 (p<0.001) 34 -7.7% 35 Mixed practices
36 1408 1409 1400 1399 1393 1378 1370 1357 1341 1314 1299 (p<0.001) 37 38 Total number of -3.9%
39 GPs 6247 6251 6259 6252 6216 6189 6165 6135 6101 6055 6003 (p<0.001) 40 41 Total number of -1.2%
42 general practices 6427 6415 6417 6408 6395 6382 6383 6372 6370 6356 6350 (p<0.001) 43 Practices with a 44 vacant GP post and Proportion of practices with a 45 their distribution by vacant GP post [95%CI] 46 type of practice (%) 47 48 1.36% 2.62%
49 Adult practices 45 44 38 38 46 41 53 58 63 64 86 [0.96-1.75] [2.08-3.17]
50 (25.0%) (24.8%) 51 52 1.44% 4.98%
Paediatric practices 22 16 20 19 30 33 38 46 60 65 71 53 [0.84-2.04] [3.85-6.11]
54 (12.2%) (20.5%) 55
56 8.03% 14.63%
Mixed practices 99 57 113 104 100 103 119 127 133 146 172 190 [6.61-9.44] [12.7-16.55] (54.7%) 58 (62.8%) 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 23
1 2 3 Total number of 2.8% 5.46% 4 practices with a 180 164 158 156 179 193 218 237 269 301 347 [2.4-3.2] [4.91-6.02] 5 vacant GP post 6 Number of 7 practices with a 1.26% 3.78% 8 81 93 83 86 93 9 vacant GP post for 108 131 142 160 185 240 [0.99-1.53] [3.31-4.25] 10 more than one year 11 Average age of GPs Change 12 by type of practice (p for trend) 13 3.7 yrs 14 Adult practices 53.3 53.6 53.9 54.1 54.6 55.0 55.2 55.7 56.3 56.6 57.0 (p<0.001) 15 16 5.4 yrs Paediatric practices 17 54.1 54.4 55.0 55.6 56.1 56.7 57.1 57.9 58.5 58.9 59.5 (p<0.001) 18 For peer review only 19 4.2 yrs Mixed practices 20 52.2 52.6 52.8 53.1 53.6 53.7 54.2 54.7 55.3 55.7 56.4 (p<0.001) 21 Sex distribution 22 Change among GPs 23 (p for trend) 24 (number) 25 Number of female 3.3%
26 GPs 3265 3291 3317 3344 3349 3359 3377 3383 3390 3383 3373 (p<0.001) 27 Number of male -11.8% 28 29 GPs 2982 2960 2942 2908 2867 2830 2788 2752 2711 2672 2630 (p<0.001) 30 Sex distribution Proportion of sex distribution 31 among GPs (%) [95%CI] 32 33 Representation of 52.27% 56.19% 34 female GPs 52.3 52.6 53.0 53.5 53.9 54.3 54.8 55.1 55.6 55.9 56.2 [51.03-53.50] [54.93-57.44] 35 Representation of 47.73% 43.81% 36 male GPs 47.7 47.4 47.0 46.5 46.1 45.7 45.2 44.9 44.4 44.1 43.8 [46.5-48.97] [42.56-45.07] 37 38 39 40 Table 1: Number and average age of GPs and the number of practices with a 41 42 vacant GP post by the type of practices, as well as, the sex distribution of GPs in 43 Hungary, 2007-2016 44 45 46 47 The number of GPs in all types of practices decreased during the decade, especially 48 the number of GPs in mixed (by 7.7%) and paediatric (by 6.5%) practices. Despite the 49 50 administrative measures resulting in a reduction of the number of practices, the 51 number of practices with a vacant GP post increased dramatically, especially in the 52 53 second half of the period examined. At the beginning of the decade, the number of 54 practices with unfilled GP posts was 180, while at the end of the period it was 347; 55 56 these shifts occurred across 519 settlements. By the end of the decade, the number of 57 adult practices with unfilled GP posts doubled, while the number of paediatric 58 practices with a vacancy for a paediatrician more than tripled. Although the relative 59 60 change in the case of vacant mixed practices was less (68.1% increase), at the end of
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 23 BMJ Open
1 2 3 the decade among the practices with a vacant GP post this type of practice was 4 5 dominated (adult: 24.8%, paediatric: 20.5%, mixed: 54.7%), as at the beginning of the 6 period (adult: 25.0%, paediatric: 12.2%, mixed: 62.8%). The most serious findings 7 8 were that 69.2% of the practices with a vacancy for GP in 2017 had had the vacancy 9 for more than a year, and that number was three times higher than it was in 2007 (240 10 vs 81). 11 12 13 Change in age and sex distribution of GPs 14 15 The average age of GPs was high even at the beginning of the decade, especially the 16 age of the paediatricians (54.1 yrs), but the average age of GPs serving adult and 17 18 mixed practices wasFor also overpeer fifty yearsreview (53.3 and 52.2only years, respectively). During 19 the decade examined, the average age of GPs was increased by 3.7 years in adults, by 20 21 5.4 yrs in paediatric and by 4.2 yrs in mixed practices (Table 1). Concerning the 22 balance by sex, in 2007, 52.3% of the GPs was female, and this rate increased to 23 24 56.2% by the end of the decade. The age structure graph showing the distribution of 25 female and male GPs by age in 2007 and ten years later in 2017 (Figure 1) clearly 26 indicates both ageing and feminization of the GPs population. 27 28 29 Association between deprivation and practices with an unfilled GP post 30 31 Mapping the DI values shows that the most-deprived areas were found in the north- 32 eastern and south-western parts of Hungary. The least-deprived sections were areas in 33 34 the north-western part of the country and in the capital city of Budapest and its 35 neighbouring areas (Figure 2A). The areas of with the highest relative vacancy rates 36 37 were found along the north-eastern border of Hungary and in the mid-eastern and 38 south-western parts of the country (Figure 2B). Maps for DI and practices with vacant 39 40 GP posts show remarkable overlap; the resultant pattern of excess vacancy rates 41 showed a correlation with a spatial pattern of deprivation (Figure 2C), as is verified by 42 the results of association analysis for DI quintiles and practices with vacant GP posts. 43 44 The results of the risk analysis showed an exponential association between relative 45 vacancy rate and deprivation (χ2 homogeneity = 232.18, P<0.001, χ2 linearity = 46 47 168.87, P<0.001). 48 49 Based on DI values, in the areas of highest deprivation (quintile V), nearly a fifth 50 (18.31% [15.83 - 21.16]) of the practices had a vacancy for a GP; this rate is nearly 51 52 triple that of the Hungarian average vacancy rate (6.39% [5.88 - 6.94]). Even in 53 quintile IV of the municipalities, 8.21% of the practices had vacant GP posts, i.e., the 54 55 vacancy rate exceeded the national average by 29%. In the least-deprived areas, the 56 vacancy rate was 4.07%, which was only 64% of the national average. The vacancy 57 58 rates between the areas of deprivation quintiles, I, II and III did not differ significantly 59 (Table 2). 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 23
1 2 3 Total number of Deprivation Vacancy rate (%) Relative vacancy rate 4 settlements with vacant 5 quintiles [95%CI] [95%CI] 6 GP post, 2017 D I 7 51 4.07 [3.05 - 5.32] 0.64 [0.47 - 0.84] 8 (least-deprived area) 9 10 D II 84 3.95 [3.17 - 4.88] 0.62 [0.49 - 0.77] 11 12 D III 107 4.52 [3.76 - 5.44] 0.71 [0.59 - 0.85] 13 14 15 D IV 128 8.21 [6.96 - 9.69] 1.29 [1.08 - 1.53] 16 D V 17 149 18.31 [15.83 - 21.16] 2.86 [2.44 - 3.36] 18 (most-deprived area) For peer review only 19 20 Hungary 519 6.39 [5.88 - 6.94] 1 21 22 23 Table 2: Distribution of settlements with vacant GP posts and the vacancy rates 24 compared to the national average by DI quintiles on 1 January 2017 25 26 27 28 Moving GPs to another primary care practice 29 30 31 In the period examined, a definitive movement of GPs from the more deprived 32 municipalities to the least-deprived ones could be detected (Table 3). In the least- 33 deprived (DI quintile I) areas, the number of GPs increased by 2.2%, while in the 34 35 most-deprived areas (DI V) it decreased by 8.4%. Municipalities in the DI IV and V 36 areas were losing GPs while those in DI I, II and III were gaining them. The resultant 37 38 outcome of this trend is that the distribution of GPs, as well as the access to primary 39 care services by DI quintiles, became more inequitable (Table 3). 40 41 Change between 42 Deprivation Total number Number of Number of Gain/loss in 2007 and 2017 43 quintiles of GPs, 2017 outgoing GPs incoming GPs number of GPs 44 [95%CI] 45 D I 809 53 71 18 2.22% [1.21-3.24] 46 47 D II 1802 98 118 20 1.11% [0.63-1.59] 48 49 D III 2029 120 142 22 1.08% [0.63-1.53] 50 51 D IV 1259 123 101 -22 -1.75% [-1.02- -2.47] 52 53 D V 451 93 55 -38 -8.43% [-5.86- -10.99] 54 55 Total 6350 487 487 56 57 58 Table 3: Migration of general practitioners among areas with different 59 60 deprivation status (DI-DV) between 1 January 2007 and 1 January 2017
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 23 BMJ Open
1 2 3 4 5 6 Change in rate of population underserved 7 8 Parallel with the increasing vacancy rate, the rate of people with no primary care 9 service provision by their own GP was also increasing during the period examined in 10 strong association with the deprivation. The rate of people (number of persons/100,000 11 12 people) with no primary care service provision was increasing in time, especially in 13 the most deprived quintile (Table 4). 14 15 16 17 Underserved population per 100,000 people [95%CI] 18 Year For peer review only 19 DI DII DIII DIV DV 20 21 581.09 1766.01 1724.06 3480.63 6075.86 2007 22 [571.21-591.11] [1748.4-1783.76] [1707.86-1740.38] [3453.58-3507.83] [6035.00-6116.94] 23 687.24 1465.92 1469.49 2690.60 6296.72 2008 24 [676.53-698.07] [1449.89-1482.09] [1454.53-1484.56] [2666.76-2714.60] [6254.98-6338.68] 25 1066.93 1377.62 1258.64 2979.71 4893.96 2009 26 [1053.62-1080.36] [1362.07-1393.31] [1244.78-1272.61] [2954.54-3005.03] [4857.04-4931.09] 27 1454.05 1205.07 1338.9 2628.33 5023.85 28 2010 29 [1438.55-1469.68] [1190.51-1219.77] [1324.58-1353.33] [2604.62-2652.2] [4986.32-5061.59] 1029.26 2160.00 1355.47 2931.45 5379.83 30 2011 31 [1016.25-1042.40] [2140.46-2179.66] [1341.05-1370.02] [2906.32-2956.75] [5340.88-5419.00] 1348.85 1779.92 1483.27 2929.17 6028.88 32 2012 33 [1333.96-1363.86] [1762.16-1797.80] [1468.15-1498.50] [2903.97-2954.54] [5987.54-6070.44] 34 1466.16 2059.88 1723.64 4174.78 6535.45 2013 35 [1450.69-1481.75] [2040.76-2079.14] [1707.32-1740.09] [4144.47-4205.26] [6491.99-6579.14] 36 1520.52 2173.38 1843.66 5012.76 6174.78 2014 37 [1504.78-1536.37] [2153.71-2193.19] [1826.76-1860.68] [4979.45-5046.23] [6132.51-6217.26] 38 1397.14 2160.64 2339.49 5444.36 7303.19 39 2015 40 [1382.05-1412.36] [2141.00-2180.41] [2320.44-2358.66] [5409.71-5479.19] [7257.56-7349.03] 2145.78 2251.45 3072.88 6116.59 7186.53 41 2016 42 [2127.10-2164.59] [2231.38-2271.66] [3051.00-3094.88] [6079.75-6153.61] [7141.19-7232.08] 1948.73 3331.61 3278.79 7057.39 8964.66 43 2017 44 [1930.92-1966.66] [3307.18-3356.18] [3256.18-3301.51] [7017.81-7097.14] [8914.02-9015.51] 45 46 Table 4: Rate of underserved population by year and deprivation quintile 47 48 49 50 51 52 DISCUSSION 53 54 Primary care is considered to be the foundation of the health care system throughout 55 56 the world. Currently, in countries with rapidly ageing populations and a high number 57 of people living with chronic diseases, a strong and accessible base of primary care 58 59 providers is essential to the proper level of medical services as well to health 60 promotion for the whole population. There is a general agreement in the scientific
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 23
1 2 3 literature regarding a workforce crisis at the level of primary care; the shortage of 4 5 primary care physicians now calls for attention and urgency worldwide.64 6 7 In our present report, the following trends and phenomena are clearly identified as 8 9 characteristics of the Hungarian primary care workforce crisis in the decade examined: 10 11 - Increasing number of primary care practices with a vacant GP post 12 - Decreasing number of GPs, especially in mixed and paediatric practices 13 14 - Increasing average age of GPs in practice 15 - Feminization of the profession indicated by increasing representation of female 16 17 GPs 18 - Strong associationFor betweenpeer deprived review areas and only practices with unfilled posts for 19 GP 20 21 - Intense movement of GPs from the deprived areas to the least-deprived ones 22 23 It is evident that the GP shortage has a direct effect on the quality of primary care 24 25 service provision, but it is reasonable to suppose that it causes problems at the level of 26 secondary care and in access to public health services. Although the usefulness and 27 effectiveness of the gatekeeping function of GPs in reducing utilization of emergency 28 29 and hospital services is a subject of intense discussion with contradictory 30 conclusions,see reviewed in 65-66 in the CEE countries with a very unfavourable 31 32 health status of their population,67-68 it seems to be essential. In these countries 33 (except the Czech Republic and Slovakia), patients do not have direct access to 34 35 hospital consultants, but the GPs control access to secondary care.69 Although, 36 currently, more outpatient specialist services can be approached in Hungary without 37 38 referral, the traditional referral system is still the dominant one in practice. 39 Consequently, the increasing number of practices with unfilled GP posts may strongly 40 41 affect the health status of the populations not only by missing primary care and 42 preventive services, but through a backlog of referrals to relevant specialists, as well. It 43 is worth mentioning that the population of the most deprived areas is the most affected 44 45 by the GP shortage and the replacement of GPs in practices especially in small rural 46 settlements is only partial, if any. 47 48 49 The chronic shortage of doctors in the primary health care sector owing to low pay, 50 restricted opportunities for training or career progression and poor working conditions 51 has been a well-known fact for a long time,70-71 but in the last decade, an 52 53 exponential increase in vacant GP posts was observed in different countries. The Pulse 54 magazine survey of 860 GPs in May 2017 found that 12.2% of positions were 55 56 currently vacant in England, an increase from 2.1% in 2011 when Pulse started 57 collecting these data.72 Similarly, in a survey in Scotland, 22% of responding 58 59 primary care practices reported current GP vacancies as of 31 August 2015, an 60 increase from 9% of practices reporting current vacancies in 2013.73 The same
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 23 BMJ Open
1 2 3 phenomenon can be detected in Hungary, where the number of primary care practices 4 5 with an unfilled GP post almost tripled in the period of 2010-2017. The decrease is 6 significant in the number of GPs in all types of practices, but it is more pronounced in 7 8 mixed and paediatric practices. It seems that the mixed general practices are less 9 attractive among GPs. One of the possible reasons of this phenomenon is that most of 10 these practices are found in small settlements where the GP is the only primary care 11 12 service provider. The reason of the decrease in the number of paediatricians is also 13 clear; the paediatrician medical specialization training focuses mostly on the hospital 14 15 replacement in Hungary. Hospitals in many parts of the country are currently reporting 16 shortages in paediatricians and young medical graduates take the hospital work more 17 18 attractive and the carrierFor paths peer more clear review at present. only 19 20 Parallel with the decrease in the number of GPs, the average age of the GPs became 21 significantly higher. The age distribution clearly indicates that the number of male GPs 22 23 under the age of 60 years was lower in 2017 than in 2007 across all ages, while in the 24 case of female GPs, although the trend is very similar, increases can be seen at certain 25 26 ages, mainly at younger ones. This fact is in harmony with the observed feminization 27 of the profession. In a systematic review based on 32 relevant publications to examine 28 the evidence that quantifies the effect of feminization on practice characteristics, 29 30 Hedden and her colleagues74 note “that female GPs self-report fewer hours of work 31 than male GPs, have fewer patient encounters, and deliver fewer services, but spend 32 33 longer with their patients during a contact and address more separate presenting 34 problems in one visit”, and conclude that “the feminization of the workforce is likely 35 36 to have a small negative impact on the availability of primary health care services, and 37 the drivers of observed differences between male and female GPs are complex and 38 39 nuanced.” Considering the fact that Hungary is among the 10 countries that have the 40 best parental leave policies in the world ― the length of paid maternity leave is 24 41 weeks and is paid at 70% of salary, which can be extended by an additional time off 42 43 paid at 40% of salary for 136 weeks75 ― female doctors may stay away from work 44 for a fairly long time, which can make finding a suitable temporary replacement 45 46 problematic. 47 48 Behind the fact that in Hungarian primary care there is a strong association between 49 50 deprived areas and practices with unfilled posts for GPs and the intense movement of 51 GPs from the deprived areas to the least-deprived ones is the probability of an existing 52 financial background. Although the majority of GPs are private entrepreneurs 53 54 capitation-based financing is almost exclusively based on the list size of the practices, 55 and only approximately 3% of GPs’ reimbursement is pay-for-performance. 56 57 Consequently, GPs are financially motivated to maximize the number of persons in the 58 practice and minimize performance.39 According to the estimated data from the 59 60 Hungarian Medial Chamber, in 2017, after covering the expenses related to
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 23
1 2 3 maintaining the infrastructure and salaries of the other health staff (at least a nurse) the 4 5 remaining income that can be used to pay the salary of the GP was approximately 6 1000 EUR before taxation. Although it can be – and should be – refused on the basis 7 8 of ethical considerations, it remains a fact that the informal so called “gratuity” 9 payment from the patients for GP services is more or less essential to make a living. It 10 is reasonable to suppose that moving GPs from deprived areas to least-deprived ones is 11 12 ― at least in part ― motivated by the expected increase in “gratuity” payment. These 13 observations clearly indicate that departmental intervention is needed and raise the 14 15 decision-making responsibility to support effective recruitment and retention of GPs 16 not only for the most-deprived areas, but for the Hungarian primary care system as a 17 18 whole, as well. For peer review only 19 20 The recruitment and retention problems are well known in the primary care systems of 21 22 developed countries, but information about effective strategies and interventions is 23 very limited. Recently, a systematic review was undertaken in which 51 studies 24 25 assessing 42 interventions categorized into 13 groups (financial incentives, recruiting 26 rural students, international recruitment, rural- or primary-care-focused undergraduate 27 28 placements, rural or underserved postgraduate training, well-being or peer support 29 initiatives, marketing, mixed interventions, support for professional development or 30 31 research, retainer schemes, re-entry schemes, specialized recruiters or case managers 32 and delayed partnerships) were retrieved. Weak evidence supported the use of 33 postgraduate placements in underserved areas, undergraduate rural placements and 34 35 recruiting students to medical school from rural areas, and there was mixed evidence 36 about financial incentives.76 37 38 39 A single report on a longitudinal small area whole country study using lower super 40 output area data and the Index of Multiple Deprivation 2010 rankings for England 41 described substantial inequalities in GPs’ service, expressed in full-time equivalent 42 43 (FTE) GPs before the ‘Equitable Access to Primary Medical Care’ (EAPMC) 44 programme was launched in December 2007. EMPMC invested £250 million towards 45 46 establishing new general practices and GP-led health centres, as well as towards 47 extending opening hours and expanding services in the 38 most “underdoctored” 48 49 Primary Care Trust areas and providing a guaranteed competitive salaries for GPs. 50 Total numbers of FTE for GPs have grown much faster in the most-deprived fifth of 51 small areas in England than elsewhere, with the GP supply in the most affluent fifth 52 53 growing at the slowest pace over the 10 years covered by the study.16 Our findings 54 obtained in our longitudinal study strongly differ from the results presented as 55 56 outcomes of the EAMC programme. It is reasonable to suppose that the main cause of 57 the difference that competitive salaries for GPs are not guaranteed in Hungary. 58 59 According to OECD figures, the income of a GP in Denmark is five times larger than 60 the income of a GP in Hungary, and the average GP in Scandinavia earns 5-10 times
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 23 BMJ Open
1 2 3 more than the average GP in Hungary.77 In 2013, a salary increase programme for 4 5 GPs started in Hungary, but the annual growth rate in nominal terms was as low as 6 2.3%.78 However, since 2014, a single settlement allowance has been given to GPs 7 8 at the time of entering into a permanently vacant GP posts, no permanent incentives 9 were introduced to support movement by GPs into underserved deprived areas. 10 11 Although studies on the effect of financial incentives for GPs in Hungary are lacking, 12 it can be stated that without increasing the salaries of GPs there is no chance to stop 13 14 the deepening workforce crisis in Hungarian primary care. 15 16 The limitation of our study is that it is an ecological one that assumes that individual 17 members of a group all have the average characteristics of the group as whole, when, 18 For peer review only 19 in fact, any association observed between variables at the group level does not 20 necessarily mean that the same association exists for an individual plucked from the 21 22 group. In addition, we had no opportunity to study the effects of different health care 23 reforms on the primary care workforce in Hungary. 24 25 26 27 FIGURE LEGENDS 28 29 Figure 1: The age and sex distribution of GPs in 2007 and 2017 in Hungary 30 31 Figure 2: The distribution pattern of deprivation (A) and of the relative vacancy rates 32 (B) in Hungary and the relationship between deprivation and the relative vacancy rate 33 34 (C) 35 36 37 38 REFERENCES 39 40 1 Campbell J, Dussault G, Buchan J, et al. A universal truth: no health without a 41 workforce. Forum report, Third Global Forum on Human Resources for Health, 42 43 Recife, Brazil. Geneva: World Health Organization, 2013. 44 https://www.who.int/workforcealliance/knowledge/resources/GHWA_AUniversal 45 TruthReport.pdf (accessed Feb 2019) 46 2 Hays R, Pong LT, Leopando et al. Primary care in the Asia-Pacific region: 47 48 challenges and solutions. Asia Pac Fam Med 2012;11:8. doi:10.1186/1447-056X- 49 11-8 50 3 Strasser R, Kam SM, Regalado SM. Rural health care access and policy in 51 developing countries. Annu Rev Public Health. 2016;37:395-412. doi: 52 53 10.1146/annurev-publhealth-032315-021507. Epub 2016 Jan 6. 54 4 Berwick DM, Hackbarth AD. Eliminating waste in US health care. JAMA. 55 2012;14:1513–6. doi: 10.1001/jama.2012.362. 56 5 Petterson, SM, Liaw, WR, Tran, C, et al. Estimating the Residency Expansion 57 58 Required to Avoid Projected Primary Care Physician Shortages by 2035. Ann Fam 59 Med 2015;2:107–14. doi:10.1370/afm.1760 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 23
1 2 3 6 Dale J, Potter R, Owen K, et al. Retaining the general practitioner workforce in 4 5 England: what matters to GPs? A cross-sectional study. BMC Fam Pract 6 2015;1:140. doi:10.1186/s12875-015-0363-1. 7 7 Hayhoe B, Majeed A, Hamlyn M, et al. Primary care workforce crisis: how 8 many more GPs do we need? RCGP Annual Conference, Harrogate, 2016. 9 10 8 Majeed A. Shortage of general practitioners in the NHS. BMJ 2017;358:j3191. 11 doi: https://doi.org/10.1136/bmj.j3191 12 9 Burla L, Widmer M. Le corps médical en Suisse: effectif et évolutions jusqu’en 13 2011. Neuchâtel: Observatoire suisse de la santé; 2012. 14 15 https://www.obsan.admin.ch/sites/default/files/publications/2015/obsan_bulletin_2 16 012-03_f.pdf (accessed Feb 2019) 17 10 Seematter-Bagnoud L, Junod J, Jaccard Ruedin H, et al. Offre et recours aux 18 soins médicauxFor ambulatoires peer en Suisse review – Projections only à l’horizon 2030. Neuchâtel: 19 Observatoire suisse de la santé; 2008. 20 21 https://www.obsan.admin.ch/sites/default/files/publications/2015/arbeitsdokument- 22 33.pdf (accessed Feb 2019) 23 11 OECD/EU (2016), Health at a Glance: Europe 2016: State of Health in the EU 24 Cycle. OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264265592-en 25 26 12 Comino EJ, Davies GP, Krastev Y, et al. A systematic review of interventions 27 to enhance access to best practice primary health care for chronic disease 28 management, prevention and episodic care. BMC Health Serv Res 2012;12:415 doi: 29 10.1186/1472-6963-12-415 30 31 13 IMO Position Paper on Health Inequalities, Irish Medical Association. 2012. 32 https://www.imo.ie/policy-international-affair/overview/IMO-Position-Paper-on- 33 Health-Inequalities.pdf (accessed Feb 2019) 34 35 14 Bauer J, Brueggmann D, Ohlendorf D, et al. General practitioners in German 36 metropolitan areas - distribution patterns and their relationship with area level 37 measures of the socioeconomic status. BMC Health Serv Res. 2016 Nov 38 25;16(1):672. doi: 10.1186/s12913-016-1921-5 39 40 15 Baker M, Ware J, Morgan K. Time to put patients first by investing in general 41 practice. Br J Gen Pract 2014;64:268–9. doi:10.3399/bjgp14X679921 42 16 Asaria M, Cookson R, Fleetcroft R, et al. Unequal socioeconomic distribution 43 of the primary care workforce: whole-population small area longitudinal study. 44 45 BMJ Open 2016;1:e008783. doi: 10.1136/bmjopen-2015-008783 46 17 Steinhaeuser J, Otto P, Goetz K, et al. Rural area in a European country from a 47 health care point of view: an adaption of the Rural Ranking Scale. BMC Health 48 Serv Res. 2014;14:147. doi: 10.1186/1472-6963-14-147. 49 50 18 Kringos D, Boerma W, Hutchinson A, et al. Building primary care in a 51 changing Europe. 38th ed. Copenhagen; 2015 52 http://www.euro.who.int/__data/assets/pdf_file/0018/271170/BuildingPrimaryC 53 (accessed Feb 2019) 54 areChangingEurope.pdf 55 19 Yardim M, Uner S. Geographical disparities in the distribution of physicians in 56 Turkey. TAF Prev Med Bull 2013;12:487–94. doi: 10.5455/pmb.1-1351692762 57 20 Matsumoto M, Inoue K, Farmer J, et al. Geographic distribution of primary care 58 59 physicians in Japan and Britain. Health Place 2010;16:164–6. doi: 60 10.1016/j.healthplace.2009.07.005
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 23 BMJ Open
1 2 3 21 Petterson SM, Phillips RL Jr, Bazemore AW, et al. Unequal distribution of the 4 5 US primary care workforce. Am Fam Physician 2013;87;11, pp. 6 http://www.aafp.org/afp/2013/0601/od1.pdf (accessed Feb 2019) 7 22 Hann M, Gravelle H. The maldistribution of general practitioners in England 8 and Wales: 1974–2003. Br Gen Pract 2004;54:894–8. 9 10 23 Isabel C, Paula V. Geographic distribution of physicians in Portugal. Eur J 11 Health Econ 2010;11:383–93. doi: 10.1007/s10198-009-0208-8 12 24 Fülöp G, Kopetsch T, Hofstätter G, et al. Regional distribution effects of ‘needs 13 planning’ for office-based physicians in Germany and Austria-methods and 14 15 empirical findings. J Public Health 2008;16:447–55. doi: 10.1007/s10389-008- 16 0187-8 17 25 Brown M. Using Gini-style indices to evaluate the spatial patterns of health 18 practitioners: TheoreticalFor peer considerations review and an application only based on Alberta data. 19 20 Soc Sci Med 1994;38:1243–56. doi: 10.1016/0277-9536(94)90189-9 21 26 Stapleton G, Schröder-Bäck P, Brand H, et al. Health inequalities and regional 22 specific scarcity in primary care physicians: ethical issues and criteria. Int J Public 23 Health 2014;59:449–55. doi: 10.1007/s00038-013-0497-7 24 25 27 World Health Organization, The World Health Report - Primary health care 26 (now more than ever). World Health Organization, 2008. 27 https://www.who.int/whr/2008/en/ (accessed Feb 2019) 28 28 World Health Organization, Increasing access to health workers in remote and 29 30 rural areas through improved retention. Background paper. Geneva, 2009. 31 https://www.who.int/hrh/retention/guidelines/en/ (accessed Feb 2019) 32 29 Watson J, Humphrey A, Peters Klimm F, et al. Motivation and satisfaction in 33 GP training: a UK cross-sectional survey. Br J Gen Pract 2011; 591: e645–9. doi: 34 35 10.3399/bjgp11X601352 36 30 Fletcher E, Abel GA, Anderson R, et al. Quitting patient care and career break 37 intentions among general practitioners in South West England: findings of a census 38 survey of general practitioners. BMJ Open 2017, doi: 10.1136/bmjopen-2017- 39 40 015853 41 31 Blane DN, McLean G, Watt G. Distribution of GPs in Scotland by age, gender 42 and deprivation. Scott Med J 2015;4:214-9. doi:10.1177/0036933015606592. Epub 43 2015 Sep 23. 44 45 32 Iacobucci G. GPs in Northern Ireland face "full blown crisis," BMA warns. 46 BMJ. 2016 Jun 7;353:i3202. doi: 10.1136/bmj.i3202 47 33 Zieler P. Physician shortage in Saxony-Anhalt. The former youngster returns as 48 country doctor. MMW Fortschr Med. 2015;17:36-7. 49 50 34 Schmidt S, Gresser U. Development and consequences of physician shortages 51 in Bavaria. Versicherungsmedizin. 2014;1:25-9. 52 35 Roeger LS, Reed RL, Smith BP. Equity of access in the spatial distribution of 53 GPs within an Australian metropolitan city. Aust J Prim Health. 2010;4:284-90. 54 55 doi: 10.1071/PY10021. 56 36 Ádány R, Kósa K, Sándor J, et al. General practitioners' cluster: a model to reorient 57 primary health care to public health services. Eur J Public Health. 2013; 23:529-530. 58 doi: 10.1093/eurpub/ckt095. 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 23
1 2 3 37 Sándor J, Kósa K, Fürjes G, et al. Public health services provided in the framework 4 5 of general practitioners' clusters. Eur J Public Health. 2013; 23:530-532. doi: 6 10.1093/eurpub/ckt096. 7 38 Kósa K, Sándor J, Dobos É, et al. Human resources development for the operation 8 of general practitioners' cluster. Eur J Public Health. 2013; 23:532-533. doi: 9 10 10.1093/eurpub/ckt097. 11 39 Sándor J, Kósa K, Papp M, et al. Capitation-Based Financing Hampers the 12 Provision of Preventive Services in Primary Health Care. Front Public Health 13 2016;4:200, doi: 10.3389/fpubh.2016.00200. eCollection 2016 14 15 40 Lember M, Cartier T, Bourgueil Y. Structure and organization of primary care. 16 In: Kringos DS, Boerma WGW, Hutchinson A (eds), Building primary care in a 17 changing Europe, Observatory Studies Series, No. 38. 2015, p. 41-9. 18 41 Juhasz A, NagyFor C, Paldypeer A. Development review of a deprivationonly index and its relation 19 to premature mortality due to diseases of the circulatory system in Hungary, 1998- 20 21 2004. Soc Sci Med 2010;70:1342–9. doi: 10.1016/j.socscimed.2010.01.024. 22 42 Nagy C, Juhasz A, Beale L, et al. Mortality amenable to health care and its 23 relation to socio-economic status in Hungary, 2004-08. Eur J Public Health 24 2012;22:620–4. doi: 10.1093/eurpub/ckr143 25 26 43 Nagy C, Juhasz A, Papp Z, et al. Hierarchical spatio-temporal mapping of 27 premature mortality due to alcoholic liver disease in Hungary, 2005-2010. Eur J 28 Public Health 2014;24:827–833. doi: 10.1093/eurpub/ckt169 29 44 Boruzs K, Juhasz A, Nagy C, et al. Relationship between statin utilization and 30 31 socioeconomic deprivation in Hungary. Front Pharmacol 2016;7:66. doi: 32 10.3389/fphar.2016.00066 33 45 Jakab Zs, Juhasz A, Nagy C, et al. Trends and territorial inequalities of 34 incidence and survival of childhood leukaemia and their relations to socioeconomic 35 36 status in Hungary, 1971–2015. Eur J Cancer Prev 1. 2017. doi: 37 10.1097/CEJ.0000000000000386. 38 46 Fairburn J, Maier W, Braubach M. Incorporating Environmental Justice into 39 Second Generation Indices of Multiple Deprivation: Lessons from the UK and 40 41 Progress Internationally. Int J Environ Res Public Health 2016;13:750. 42 doi:10.3390/ijerph13080750 43 47 Department for Communities and Local Government: English Indices of 44 Deprivation 2015. http://www.gov.uk/government/statistics/english-indices-of- 45 46 deprivation-2015 (accessed Feb 2019) 47 48 The Scottish Government Scottish Index of Multiple Deprivation. 48 http://www.gov.scot/Topics/Statistics/SIMD (accessed Feb 2019) 49 50 49 Welsh Government Welsh Index of Multiple Deprivation (WIMD). 51 http://gov.wales/statistics-and-research/welsh-index-multiple- 52 deprivation/?lang=en (accessed Feb 2019) 53 54 50 Northern Ireland Multiple Deprivation Measure 2017 (NIMDM2017). 55 https://www.nisra.gov.uk/statistics/deprivation/northern-ireland-multiple- 56 deprivation-measure-2017-nimdm2017 (accessed Feb 2019) 57 51 Havard S, Deguen S, Bodin J, et al. A small-area index of socioeconomic 58 59 deprivation to capture health inequalities in France. Soc Sci Med 2008;67:2007–16. 60 doi:10.1016/j.socscimed.2008.09.031
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 23 BMJ Open
1 2 3 52 Sánchez-Cantalejo C, Ocana-Riola R, Fernández-Ajuria A. Deprivation index 4 5 for small areas in Spain. Soc Indic Res 2008;89:259–273. doi: 10.1007/s11205- 6 007-9114-6 7 53 Domínguez-Berjón MF, Borrell C, Cano-Serral G, et al. Constructing a 8 deprivation index based on census data in large Spanish cities(the MEDEA 9 10 project). Gac Sanit 2008;22:179–87. doi:https://doi.org/10.1157/13123961 11 54 Kroll LE, Schumann M, Hoebel J, et al. Regional health differences – 12 developing a socioeconomic deprivation index for Germany. Journal of Health 13 Monitoring Published Online First: 2017. doi: 10.17886/RKI-GBE-2017-048 14 15 55 Meijer M, Engholm G, Grittner U, et al. A socioeconomic deprivation index for 16 small areas in Denmark. Scand J Public Health 2013;41:560–9. 17 doi:10.1177/1403494813483937 18 56 Panczak R,For Galobardes peer B, Voorpostel review M, et al . onlyA Swiss neighbourhood index 19 of socioeconomic position: development and association with mortality. J 20 21 Epidemiol Community Health 2012;66:1129. doi:10.1136/jech-2011-200699 22 57 Caranci N, Biggeri A, Grisotto L, et al. L’indice di deprivazione italiano a 23 livello di sezione di censimento: definizione, descrizione e associazione con la 24 mortalità. Epidemiol Prev 2010;34:167–76. 25 26 58 Šlachtová H, Tomášková H, Šplíchalová A, et al. Czech socio-economic 27 deprivation index and its correlation with mortality data. International Journal of 28 Public Health 2009;54:267–273. 29 59 Zadnik V, Guillaume E, Lokar K, et al. Slovenian Version of The European 30 31 Deprivation Index at Municipal Level. Zdr Varst 2018;57:47–54. doi:10.2478/sjph- 32 2018-0007 33 60 Besag J, York J, and Mollié A, A Bayesian image restoration with two 34 applications in spatial statistics. Ann Inst Stat Math.1991;43:1–20. doi: 35 36 10.1007/BF00116466 37 61 R Development Core Team. R: A Language and Environment for Statistical 38 Computing, R Foundation for Statistical Computing, Vienna, Austria, 2017. 39 (accessed Feb 2019) 40 www.R-project.org 41 62 Rue H, Martino S, Chopin N. Approximate Bayesian inference for latent 42 Gaussian models using integrated nested Laplace approximations (with discussion). 43 Journal of the Royal Statistical Society, Series B, 2009;2:319-392, www.r-inla.org 44 45 (accessed Feb 2019) 46 63 Beale L, Hodgson S, Abellan JJ, et al. Evaluation of spatial relationships 47 between health and the environment: the rapid inquiry facility. Environ Health 48 Perspect 2010;118, 1306–12. doi: 10.1289/ehp.0901849 49 50 64 Song Z, Chopra V, McMahon LF. Addressing the Primary Care Workforce 51 Crisis. Am J Manag Care 2015;8:e452-4. 52 65 Van den Heede K, Van de Voorde C. Interventions to reduce emergency 53 department utilisation: A review of reviews. Health Policy 2016;120:1337-1349. 54 55 doi: 10.1016/j.healthpol.2016.10.002. 56 66 Greenfield G, Foley K, Majeed A. Rethinking primary care's gatekeeper role. 57 BMJ 2016;i4803. doi:10.1136/bmj.i4803 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 23
1 2 3 67 Karanikolos M, Adany R, McKee M. The epidemiological transition in Eastern 4 5 and Western Europe: a historic natural experiment. Eur J Public Health 6 2017;suppl_4:4-8. doi: 10.1093/eurpub/ckx158 7 68 Mackenbach JP. Nordic paradox, Southern miracle, Eastern disaster: 8 persistence of inequalities in mortality in Europe. Eur J Public Health 9 10 2017;suppl_4:14-17. doi: 10.1093/eurpub/ckx160 11 69 Euro Health Consumer Index 2016, Health Consumer Powerhouse Ltd, 2017. 12 https://healthpowerhouse.com/media/EHCI-2016/EHCI-2016-report.pdf (accessed 13 Feb 2019) 14 15 70 Rechel B, Dubois C-A, McKee M (eds). The Health Care Workforce in Europe 16 Learning from experience. European Observatory on Health Systems and Policies, 17 2006. http://www.euro.who.int/__data/assets/pdf_file/0008/91475/E89156.pdf 18 (accessed Feb 2019)For peer review only 19 20 71 Kuehn BM. Reports Warn of Primary Care Shortages. JAMA. 2008;300:1872– 21 5. doi: 10.1001/jama.300.16.1872. 22 72 NHSProvider. Summary of vacancy, shortfall and fill rate data for the clinical 23 workforce. November 2017. https://nhsproviders.org/media/3903/summary-of- 24 25 vacancy-shortfall-and-fill-rate-data-for-the-clinical-workforce.pdf (accessed Feb 26 2019) 27 73 Primary Care Workforce Survey Scotland 2015, A Survey of Scottish General 28 Practices and General Practice Out of Hours Services. National Services Scotland, 29 30 Information Services Division, Publication Report, 14 June, 2016 31 http://www.isdscotland.org/Health-Topics/General-Practice/Publications/2016-06- 32 14/2016-06-14-PrimaryCareWorkforceSurveyScotland2015-Report.pdf (accessed 33 Feb 2019) 34 35 74 Heeden L, Barer ML, Cardiff K, et al. The implications of the feminization of 36 the primary care physician workforce on service supply: a systematic review. 37 Human Resources for Health 2014;12:32. https://doi.org/10.1186/1478-4491-12-32 38 75 Weller C.These 10 countries have the best parental leave policies in the world, 39 40 World Economic Forum, 2016. 41 https://www.weforum.org/agenda/2016/08/these-10-countries-have-the-best- 42 parental-leave-policies-in-the-world (accessed 28 Feb 2019) 43 76 Verma P, Ford JA, Stuart A, et al. A systematic review of strategies to recruit 44 45 and retain primary care doctors. BMC Health Serv Res 2016;16:126. 46 https://doi.org/10.1186/s12913-016-1370-1 47 77 OECD Health Statistics 2017. 48 78 OECD Health Statistics 2015. 49 50 51 52 53 54 FOOTNOTES 55 56 Contributors MP, JS, CN, AJ and RA developed the study design. LK identified the 57 58 practices by location, vacancy and characteristics of GPs. MP, AJ and CN did the data 59 processing and analyses. MP wrote the first draft of the paper and RA finalized it. All 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 23 BMJ Open
1 2 3 authors critically revised the first draft and approved the final manuscript submitted for 4 5 publication. 6 Funding This work was supported by the Swiss Contribution SH/8/1 project and by 7 the GINOP-2.3.2-15-2016-00005 project financed by the European Union under the 8 9 European Social Fund and European Regional Development Fund, as well as by the 10 Hungarian Academy of Sciences (TK2016-78). 11 12 Competing interests None declared. 13 14 15 Patient consent Not required. 16 Ethics approval Not required. 17 18 Provenance and peerFor review peer Not commissioned; review externally only peer reviewed. 19 20 Data sharing statement Data were provided by the National Institute of Health 21 Insurance Fund Management (NIHIFM), the Regional Informational System of the 22 Ministry of Local Government and Regional Development, the Hungarian Central 23 24 Statistical Office and the Hungarian Tax and Financial Control Administration. Data 25 requests should be sent to Professor Roza Adany (corresponding author, 26 [email protected]). 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 170x175mm (300 x 300 DPI) 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 23 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 170x250mm (300 x 300 DPI) 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open
Workforce crisis in primary health care worldwide: the Hungarian example in a longitudinal follow-up study
Journal: BMJ Open ManuscriptFor ID peerbmjopen-2018-024957.R2 review only Article Type: Research
Date Submitted by the 01-Jul-2019 Author:
Complete List of Authors: Papp, Magor; National Public Health Institute Kőrösi, László; National Institute of Health Insurance Fund Management Sándor, János; Faculty of Public Health, University of Debrecen, Department of Preventive Medicine Nagy, Csilla; Public Health Administration Service of Government Office of Capital City Budapest Juhász, Attila; Public Health Administration Service of Government Office of Capital City Budapest Ádány, Róza; University of Debrecen, Faculty of Public Health, Department of Preventive Medicine; University of Debrecen, Debrecen, Hungary, MTA-DE Public Health Research Group of the Hungarian Academy of Sciences
Primary Subject General practice / Family practice Heading:
Secondary Subject Heading: Public health, Health services research
PRIMARY CARE, workforce crisis, vacancy, general practitioners, Keywords: socioeconomic status, deprivation
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 23 BMJ Open
1 2 3 4 Workforce crisis in primary health care worldwide: 5 the Hungarian example in a longitudinal follow-up study 6 7 8 Magor Papp1, László Kőrösi2, János Sándor3, Csilla Nagy4, Attila Juhász4, Róza Ádány3,5 9 10 1 National Public Health Institute, Budapest, Hungary 11 12 2 National Institute of Health Insurance Fund Management, Budapest, Hungary 13 14 3 Department of Preventive Medicine, Faculty of Public Health, University of Debrecen, Debrecen, Hungary 15 16 4 Public Health Administration Service of Government Office of Capital City Budapest, Budapest, Hungary 17 18 5 MTA-DE Public HealthFor Research peer Group of the reviewHungarian Academy onlyof Sciences, University of Debrecen, Debrecen, Hungary 19 20 21 Corresponding author: Róza Ádány, MTA-DE Public Health Research Group, 22 23 Department of Preventive Medicine, Faculty of Public Health, University of Debrecen, 24 25 Hungary 26 Debrecen, Kassai str 26, 27 28 H-4028 29 e-mail: [email protected] 30 Phone: +3652512764 31 32 Fax: +3652417267 33 34 35 Word count: 4513 36 37 Keywords: primary care, workforce crisis, vacancy, general practitioners, 38 39 socioeconomic status, deprivation 40 41 42 ABSTRACT 43 44 45 Objective The study was designed to explore the development of the general 46 practitioner (GP) shortage in primary care and its characteristics in Hungary. 47 48 Design Longitudinal follow-up study over the decade 2007-2016. 49 50 51 Methods Analyses were performed on changes in number, age and sex of GPs by 52 practice type (adult, paediatric and mixed), as well as on their geographical 53 distribution and migration between areas characterized by deprivation index (DI) at 54 55 municipality level. The association between deprivation and vacancy for GPs was 56 studied by risk analysis. The number of population underserved was defined by DI 57 quintile. 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 23
1 2 3 Setting and subjects The study involved all general practices and GPs in the period 4 5 examined. 6 7 Main outcome measure It is showed that the number of general practices with 8 unfilled GP posts was increasing exponentially, mainly in the most-deprived areas of 9 10 the country. 11 12 Results A decrease in the number of GPs in all types of practices, especially in mixed 13 14 (by 7.7%; p<0.001) and paediatric (by 6.5%; p<0.001) ones was shown; the number of 15 adult practices with unfilled GP posts doubled, while the number of paediatric 16 practices with a vacancy for a paediatrician more than tripled. The average age of GPs 17 18 was increased by For3.7 years peer (p<0.001) review in adult, by 5.4 only yrs (p<0.001) in paediatric and 19 by 4.2 yrs (p<0.001) in mixed practices. In 2007, 52.27% (95%CI 51.03-53.5) of the 20 21 GPs was female, and this rate increased to 56.19% (95%CI 54.93-57.44) by the end of 22 the decade. An exponential association between relative vacancy rate and deprivation 23 24 was confirmed. As a result of the migration of GPs, in the most-deprived areas the 25 number of GPs decreased by 8.43% (95%CI 5.86-10.99). 26 27 Conclusions The workforce crisis in Hungarian primary care is progressively 28 29 deepening and resulting in more severe inequity in access to health care. 30 31 32 STRENGTHS AND LIMITATIONS 33 34 35 - The study is based on a comprehensive longitudinal follow-up survey at a 36 national level to evaluate major changes in the number and distribution of 37 general practitioners (GPs). 38 - The survey used an area-based composite indicator to study the relationship 39 40 between deprivation and shortage in GPs in the country. 41 - The study analysed the migration of GPs between practices, but the reasons for 42 leaving primary care were not investigated in detail. 43 - Data were available for the first day of each year; thus the study could not 44 45 detect transient changes in a year. 46 47 48 INTRODUCTION 49 50 51 The workforce crisis in primary care (PC) is a worldwide phenomenon. The shortage 52 in general practitioners (GPs) is particularly severe in low- and middle-income 53 54 countries in Africa, Asia and the Pacific,1-3 but it is becoming more striking in high- 55 income countries as in the USA,4-5 UK,6-8 Switzerland,9-10 as well. In 2016, 56 57 England was already approximately 6,500 GPs below the ideal number, and this gap 58 would increase to 12,100 by 2020.6 Switzerland is also facing an impending primary 59 60 care workforce crisis since almost half of all primary care physicians are expected to
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 23 BMJ Open
1 2 3 retire in the next decade.9-10 As the “Health at a glance, 2016” report11 shows in 4 5 the Organisation for Economic Co-operation and Development (OECD) countries ― 6 among them in the EU member states ― the workforce crisis at the level of primary 7 8 care is so deep that in the period of 2011-2013 a significant portion of the emergency 9 department (ED) visits occurred because primary care was not available. In certain 10 11 Central Eastern European (CEE) countries (Czech Republic, Slovakia), this proportion 12 of ED visits was as high as 52-74%. 13 14 The shortage of GPs seems to be the key issue of the workforce crisis in primary care 15 16 throughout the world which may strongly contribute to the limited access of people 17 living in socioeconomically deprived conditions to primary care and other health 18 For peer review only 19 services.12 The Position Paper of the Irish Medical Organisation on health 20 inequalities denotes that vulnerable rural and deprived urban communities have no 21 22 adequate GP cover in Ireland.13 Demographic studies identified a shortage of 23 primary care service providers located in the countryside and a relative oversupply in 24 metropolitan areas in Germany, and in German cities with a population size of more 25 26 than 500,000 the distribution patterns of GPs were significantly correlated with the 27 geodemographic index, an area measure of deprivation.14 Although the Royal 28 29 College of General Practitioners reported that the so-called “underdoctored” areas are 30 the most-deprived regions of the UK, i.e., “the general practice workforce is unevenly 31 32 spread across the country, with the fewest doctors in the most-deprived areas, 33 exacerbating health inequalities”,15 it is worth mentioning that there was a 34 35 substantial reduction in socioeconomic inequality in family physician supply 36 associated with national policy for establishing new practices in the most deprived 37 38 areas under the ‘Equitable Access to Primary Medical Care’ programme.[16] 39 40 Substantial geographical inequalities in family physician supply was demonstrated 41 between large subnational areas even in high-income countries with universal health 42 43 coverage. However, because these studies have focused on large areas they have not 44 been able to accurately describe socioeconomic inequality in primary care supply by 45 46 pinpointing primary care shortages in specific disadvantaged neighbourhoods.[17-26] 47 48 The WHO Department of Human Resources for Health provided an overview on the 49 inequitable access of populations to primary care. The underserved populations were 50 51 identified in remote and rural areas where the socio-economic environment, such as 52 working and living conditions, as well as access to education for children were 53 54 unfavourable.27-28 Data are available almost exclusively from cross-sectional rather 55 than longitudinal surveys;6, 29-31 therefore, these studies are unable to report on 56 57 previous and actual translations in the primary care system. In addition, these surveys 58 were conducted among GPs typically in a single region of a country,30, 32-35 which 59 60 may not be entirely typical of a country as a whole.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 23
1 2 3 Concerning the population size of Hungary, as well as the fact that the workforce crisis 4 5 in primary care was evolving in the last decade the country may serve as an ideal field 6 to a study targeting the characterization of the workforce crisis in its development. In 7 8 Hungary responsibilities for primary care have been centralized at national level and 9 the only source of financing is the National Health Insurance Fund. The local 10 governments are responsible for ensuring that primary care services are available to 11 12 their population, so practices are organized by them at municipality level. The size of 13 the practices, reflecting the number of clients served, varies widely (800–3000 14 15 persons/practice), and the average size is approximately 1550 persons/practice. 16 Generally, more family practices operate in more highly populated municipalities, 17 18 whereas one familyFor practitioner peer serves review more than one only municipality in less populated 19 areas. Primary care physicians are mostly (95%) self-employed service providers 20 contracted to the National Health Insurance Fund, and they work almost exclusively in 21 22 solo practices, i.e. one general practitioner - with the assistance of one practice nurse - 23 serves the clients. GPs provide a complexity of care (screening, health status 24 25 assessment, diagnosis, treatment, follow-up and referral to the secondary level of care) 26 through capitation based insurance coverage for this basic health care service. 27 28 Concerning the type of practices, about half of them serves only adults, one quarter of 29 them serves exclusively children, while the remainings are mixed practices. If a GP 30
31 post becomes vacant it is very difficult to recruit replacement GP, in general GPs from 32 the nearby practices as locum-tenents provide services in very limited working hours, 33 so the population belonging to the practice with vacant GP post will be definitively 34 35 underserved. Theoretically for GPs there are some competitive work opportunities, 36 because about 40% of them in addition to specialization in family care have medical 37 38 specialization in paediatrics or internal medicine, but the health care institutions at 39 secondary or tertiary levels rarely provide GPs with positions, so they can work almost 40 41 exclusively in the emergency service at primary care level. On the basis of the results 42 obtained in the Hungarian Public Health Focused Primary Care Model Programme[36- 43 39] the government started to disseminate the GPs’ cluster work in a new pilot 44 45 programme in 2018. GPs’ clusters are built on the cooperation of six GPs employing 46 health professionals (public health professionals, community nurses, physiotherapists, 47 48 dieticians, health psychologists) who – in addition to the traditional patient care - are 49 competent to plan and implement various public health services at cluster level. In 50 51 2019, approximately 6% of the primary care practices is involved in this pilot, i.e. solo 52 practice is still the norm in Hungary. On the basis of the three dimensions of primary 53 care structure – governance, economic conditions and workforce development – the 54 55 Hungarian primary care system is qualified into the medium category,[40] but the 56 coordination and comprehensiveness are found to be weak.[18] 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 23 BMJ Open
1 2 3 Our study is designed to follow the development of the GP shortage and to describe 4 5 the characteristics of changes in Hungary, a country which illustrates a prime example 6 of CEE countries with a workforce crisis in primary care. The secular trend by changes 7 8 in the number, age and sex of GPs by type of practices, as well as the distribution of 9 GPs and vacant GP positions by deprivation are characterized in a longitudinal follow- 10 up design in the 10-year period that began on 1 January 2007 and ended on 1 January 11 12 2017. 13 14 15 16 METHODS 17 18 Data on general practicesFor peer analysed review only 19 20 The analysis is based on data provided by the National Institute of Health Insurance 21 22 Fund Management (NIHIFM) for the period of 1 January 2007 – January 2017 in 23 Hungary. The NIHIFM provided detailed information for the first day of each year on 24 the number, type (adult, paediatric or mixed) and geographical (municipal) location of 25 26 general practices, identified by numeric practice codes. It was reported whether the GP 27 post of the practice was filled or unfilled on the first day of each year. In the case of 28 29 practices with filled GP posts, the age and sex of the GPs were registered. If the GP 30 post was unfilled, the period during which there was an inability to recruit GP, as well 31 32 as the number of population underserved (i.e. the number of people belonging to 33 practice with vacant GP post) were also indicated. If the reason for vacancy involved 34 moving a GP from one primary care practice to another, the deprivation status of the 35 36 municipalities where the former and the newly engaged practices were found was also 37 considered in the analysis; the number of practices affected was defined by deprivation 38 39 index (DI) quintiles. 40 41 In the analysis performed, the change in the number of practices with filled and 42 unfilled GP posts and the change in the average age of GPs were defined by the type 43 44 of practices for the period investigated. In addition, the change in distribution by sex 45 was also determined. 46 47 48 Deprivation 49 50 To see the relationship between the distribution of unfilled practices and the socio- 51 economical status (SES) of the settlements where the practices are located, association 52 53 analyses were carried out. The SES was characterized by the DI, an area-based 54 composite indicator at the municipality level. The DI was built from seven different 55 56 indicators (income, the level of education, the rate of unemployment, the rate of one- 57 parent families, the rate of large families, the density of housing and car ownership) 58 derived from the Regional Informational System of the Ministry of Local Government 59 60 and Regional Development. The data were originally obtained from the Hungarian
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 23
1 2 3 Central Statistical Office (from the Census 2011) and the Hungarian Tax and Financial 4 5 Control Administration (for the year of the census). The development of the DI was 6 carried out by using principal component analysis. All variables were first transformed 7 8 by using the natural log-transformation and then standardised. Areas with positive 9 (higher) index values were municipalities with a lower socio-economic status than the 10 national average, and the converse was shown in areas with negative (lower) index 11 12 values. This method to calculate DI values has been published previously by us41 13 and successfully used in several studies.42-45 Same or similar approaches measuring 14 15 SES were also applied in studies carried out in other countries all over Europe[46] 16 such as the United Kingdom,[47-50] France,[51] Spain,[52-53] Germany,[54] 17 18 Denmark,[55] Switzerland,[56]For peer Italy,[57] review Czech Republic,[58] only and Slovenia.[59] 19 20 21 22 Mapping and risk analysis of vacancy 23 24 25 A hierarchical Bayesian binomial model was used to predict the relative vacancy ratio 26 at the municipality level. The data consisted of the observed number of practices with 27 unfilled GP posts and total number of practices for each municipality. The convolution 28 29 model, originally proposed by Besag et al,60 was used and incorporates both a 30 heterogeneous and spatially structured random effect. The parameters for the spatial 31 32 model were estimated via Integrated Nested Laplace Approximations (INLA) directly 33 in R using the package INLA.61-62 The map for vacancy shows the relative vacancy 34 35 rate and exceedance posterior probabilities. 36 37 Using the risk analysis capabilities of the Rapid Inquiry Facility,63 the association 38 39 between deprivation and vacancy for GPs was calculated. Relative frequencies of 40 vacancy in relation to the national average were calculated by DI quintiles as a 41 municipality-based categorical covariate. Chi-square tests for homogeneity and for 42 43 linear trend were also carried out to test the global association of the DI and the 44 vacancy. 45 46 47 48 Patient and Public Involvement 49 50 Patients and or public were not involved. 51 52 53 54 55 RESULTS 56 57 The number of local administrative units 2 (LAU2), called municipalities or 58 59 settlements in Hungary was 3176 with a total population number decreasing from 60 10,066,158 (2007) to 9,797,561 (2017) in the period examined.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 23 BMJ Open
1 2 3 At the beginning of the decade covered by the study, there were 6427 primary care 4 5 practices serving the country’s population, and at the end of the period, 6350 practices 6 were operating. During the survey period, the number of practices decreased by 77. 7 8 Behind the change in the number of practices were decisions made by local 9 governments having the right to merge practices that typically could not be fulfilled 10 for years in small-sized settlements with other practices belonging to the same 11 12 administrative municipality. 13 14 15 16 Change in number of practices with unfilled GP posts 17 18 For peer review only 19 In Hungary, the number of GPs in the beginning of the decade examined was 6247. Of 20 these, 53.2% served adults, 24.3% served children exclusively, and the remaining 21 22 22.5% served mixed practices. For the end of the period, 6003 GPs were registered 23 with the following distribution: 54.6% in adult practices, 23.8% in paediatric practices, 24 and 21.6% in mixed practices (Table 1). 25 26 27 28 2007 2017 29 2008 2009 2010 30 31 01.01.2007 01.01 01.01 01.01 01.01.2011 01.01.2012 01.01.2013 01.01.2014 01.01.2015 01.01.2016 01.01.2017 32 Number of GPs by Change
33 type of practice (p for trend) 34 -1.1% 35 Adult practices 36 3315 3316 3339 3331 3319 3317 3307 3303 3301 3301 3279 (p=0.004) 37 -6.5% 38 Paediatric practices 39 1524 1526 1520 1522 1504 1494 1488 1475 1459 1440 1425 (p<0.001) 40 -7.7% 41 Mixed practices
1408 1409 1400 1399 1393 1378 1370 1357 1341 1314 1299 (p<0.001) 42 43 Total number of -3.9% 44
GPs 6247 6251 6259 6252 6216 6189 6165 6135 6101 6055 6003 (p<0.001) 45 46 Total number of -1.2% 47
general practices 6427 6415 6417 6408 6395 6382 6383 6372 6370 6356 6350 (p<0.001) 48 49 Practices with a 50 vacant GP post and Proportion of practices with a 51 their distribution by vacant GP post [95%CI] 52 type of practice (%) 53 54 1.36% 2.62%
Adult practices 45 44 38 38 46 41 53 58 63 64 86 55 [0.96-1.75] [2.08-3.17]
56 (25.0%) (24.8%) 57 58 1.44% 4.98%
Paediatric practices 22 16 20 19 30 33 38 46 60 65 71 59 [0.84-2.04] [3.85-6.11] 60 (12.2%) (20.5%)
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 23
1 2 3 4 8.03% 14.63%
Mixed practices 99 5 113 104 100 103 119 127 133 146 172 190 [6.61-9.44] [12.7-16.55] (54.7%) 6 (62.8%) Total number of 7 2.8% 5.46% practices with a 8 180 164 158 156 179 193 218 237 269 301 347 [2.4-3.2] [4.91-6.02] 9 vacant GP post 10 Number of 11 practices with a 1.26% 3.78% 81 93 83 86 93 12 vacant GP post for 108 131 142 160 185 240 [0.99-1.53] [3.31-4.25] 13 more than one year 14 Average age of GPs Change 15 by type of practice (p for trend) 16 3.7 yrs 17 Adult practices 18 For53.3 53.6 peer53.9 54.1 54.6 review55.0 55.2 55.7 56.3 only56.6 57.0 (p<0.001) 19 5.4 yrs 20 Paediatric practices 21 54.1 54.4 55.0 55.6 56.1 56.7 57.1 57.9 58.5 58.9 59.5 (p<0.001) 22 4.2 yrs 23 Mixed practices 24 52.2 52.6 52.8 53.1 53.6 53.7 54.2 54.7 55.3 55.7 56.4 (p<0.001) 25 Sex distribution Change 26 among GPs (p for trend) 27 (number) 28 Number of female 3.3% 29 30 GPs 3265 3291 3317 3344 3349 3359 3377 3383 3390 3383 3373 (p<0.001) 31 Number of male -11.8%
32 GPs 2982 2960 2942 2908 2867 2830 2788 2752 2711 2672 2630 (p<0.001) 33 Sex distribution Proportion of sex distribution 34 among GPs (%) [95%CI] 35 36 Representation of 52.27% 56.19%
37 female GPs 52.3 52.6 53.0 53.5 53.9 54.3 54.8 55.1 55.6 55.9 56.2 [51.03-53.50] [54.93-57.44] 38 39 Representation of 47.73% 43.81%
40 male GPs 47.7 47.4 47.0 46.5 46.1 45.7 45.2 44.9 44.4 44.1 43.8 [46.5-48.97] [42.56-45.07] 41 42 43 44 Table 1: Number and average age of GPs and the number of practices with a 45 vacant GP post by the type of practices, as well as, the sex distribution of GPs in 46 47 Hungary, 2007-2016 48 49 The number of GPs in all types of practices decreased during the decade, especially 50 the number of GPs in mixed (by 7.7%) and paediatric (by 6.5%) practices. Despite the 51 52 administrative measures resulting in a reduction of the number of practices, the 53 number of practices with a vacant GP post increased dramatically, especially in the 54 55 second half of the period examined. At the beginning of the decade, the number of 56 practices with unfilled GP posts was 180, while at the end of the period it was 347; 57 58 these shifts occurred across 519 settlements. By the end of the decade, the number of 59 adult practices with unfilled GP posts doubled, while the number of paediatric 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 23 BMJ Open
1 2 3 practices with a vacancy for a paediatrician more than tripled. Although the relative 4 5 change in the case of vacant mixed practices was less (68.1% increase), at the end of 6 the decade among the practices with a vacant GP post this type of practice was 7 8 dominated (adult: 24.8%, paediatric: 20.5%, mixed: 54.7%), as at the beginning of the 9 period (adult: 25.0%, paediatric: 12.2%, mixed: 62.8%). The most serious findings 10 were that 69.2% of the practices with a vacancy for GP in 2017 had had the vacancy 11 12 for more than a year, and that number was three times higher than it was in 2007 (240 13 vs 81). 14 15 16 Change in age and sex distribution of GPs 17 18 The average age ofFor GPs was peer high even review at the beginning only of the decade, especially the 19 age of the paediatricians (54.1 yrs), but the average age of GPs serving adult and 20 21 mixed practices was also over fifty years (53.3 and 52.2 years, respectively). During 22 the decade examined, the average age of GPs was increased by 3.7 years in adults, by 23 24 5.4 yrs in paediatric and by 4.2 yrs in mixed practices (Table 1). Concerning the 25 balance by sex, in 2007, 52.3% of the GPs was female, and this rate increased to 26 56.2% by the end of the decade. The age structure graph showing the distribution of 27 28 female and male GPs by age in 2007 and ten years later in 2017 (Figure 1) clearly 29 indicates both ageing and feminization of the GPs population. 30 31 32 Association between deprivation and practices with an unfilled GP post 33 34 Mapping the DI values shows that the most-deprived areas were found in the north- 35 eastern and south-western parts of Hungary. The least-deprived sections were areas in 36 37 the north-western part of the country and in the capital city of Budapest and its 38 neighbouring areas (Figure 2A). The areas of with the highest relative vacancy rates 39 40 were found along the north-eastern border of Hungary and in the mid-eastern and 41 south-western parts of the country (Figure 2B). Maps for DI and practices with vacant 42 GP posts show remarkable overlap; the resultant pattern of excess vacancy rates 43 44 showed a correlation with a spatial pattern of deprivation (Figure 2C), as is verified by 45 the results of association analysis for DI quintiles and practices with vacant GP posts. 46 47 The results of the risk analysis showed an exponential association between relative 48 vacancy rate and deprivation (χ2 homogeneity = 232.18, P<0.001, χ2 linearity = 49 50 168.87, P<0.001). 51 52 Based on DI values, in the areas of highest deprivation (quintile V), nearly a fifth 53 (18.31% [15.83 - 21.16]) of the practices had a vacancy for a GP; this rate is nearly 54 55 triple that of the Hungarian average vacancy rate (6.39% [5.88 - 6.94]). Even in 56 quintile IV of the municipalities, 8.21% of the practices had vacant GP posts, i.e., the 57 58 vacancy rate exceeded the national average by 29%. In the least-deprived areas, the 59 vacancy rate was 4.07%, which was only 64% of the national average. The vacancy 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 23
1 2 3 rates between the areas of deprivation quintiles, I, II and III did not differ significantly 4 5 (Table 2). 6 7 Total number of Deprivation Vacancy rate (%) Relative vacancy rate 8 settlements with vacant quintiles [95%CI] [95%CI] 9 GP post, 2017 10 D I 11 51 4.07 [3.05 - 5.32] 0.64 [0.47 - 0.84] 12 (least-deprived area) 13 14 D II 84 3.95 [3.17 - 4.88] 0.62 [0.49 - 0.77] 15 16 D III 107 4.52 [3.76 - 5.44] 0.71 [0.59 - 0.85] 17 18 D IV For peer128 review8.21 [6.96 only - 9.69] 1.29 [1.08 - 1.53] 19 20 D V 149 18.31 [15.83 - 21.16] 2.86 [2.44 - 3.36] 21 (most-deprived area) 22 23 Hungary 519 6.39 [5.88 - 6.94] 1 24 25 26 Table 2: Distribution of settlements with vacant GP posts and the vacancy rates 27 28 compared to the national average by DI quintiles on 1 January 2017 29 30 31 32 Moving GPs to another primary care practice 33 34 In the period examined, a definitive movement of GPs from the more deprived 35 36 municipalities to the least-deprived ones could be detected (Table 3). In the least- 37 deprived (DI quintile I) areas, the number of GPs increased by 2.2%, while in the 38 39 most-deprived areas (DI V) it decreased by 8.4%. Municipalities in the DI IV and V 40 areas were losing GPs while those in DI I, II and III were gaining them. The resultant 41 outcome of this trend is that the distribution of GPs, as well as the access to primary 42 43 care services by DI quintiles, became more inequitable (Table 3). 44 45 Change between Deprivation Total number Number of Number of Gain/loss in 46 2007 and 2017 quintiles of GPs, 2017 outgoing GPs incoming GPs number of GPs 47 [95%CI] 48 49 D I 809 53 71 18 2.22% [1.21-3.24] 50 51 D II 1802 98 118 20 1.11% [0.63-1.59] 52 53 D III 2029 120 142 22 1.08% [0.63-1.53] 54 55 D IV 1259 123 101 -22 -1.75% [-1.02- -2.47] 56 57 D V 451 93 55 -38 -8.43% [-5.86- -10.99] 58 59 Total 6350 487 487 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 23 BMJ Open
1 2 3 4 5 Table 3: Migration of general practitioners among areas with different 6 deprivation status (DI-DV) between 1 January 2007 and 1 January 2017 7 8 9 10 11 Change in rate of population underserved 12 13 Parallel with the increasing vacancy rate, the rate of people with no primary care 14 service provision by their own GP was also increasing during the period examined in 15 strong association with the deprivation. The rate of people (number of persons/100,000 16 17 people) with no primary care service provision was increasing in time, especially in 18 the most deprived Forquintile peer(Table 4). review only 19 20 21 22 Underserved population per 100,000 people [95%CI] 23 Year 24 25 DI DII DIII DIV DV 26 581.09 1766.01 1724.06 3480.63 6075.86 2007 27 [571.21-591.11] [1748.4-1783.76] [1707.86-1740.38] [3453.58-3507.83] [6035.00-6116.94] 28 687.24 1465.92 1469.49 2690.60 6296.72 2008 29 [676.53-698.07] [1449.89-1482.09] [1454.53-1484.56] [2666.76-2714.60] [6254.98-6338.68] 30 1066.93 1377.62 1258.64 2979.71 4893.96 31 2009 [1053.62-1080.36] [1362.07-1393.31] [1244.78-1272.61] [2954.54-3005.03] [4857.04-4931.09] 32 1454.05 1205.07 1338.9 2628.33 5023.85 33 2010 34 [1438.55-1469.68] [1190.51-1219.77] [1324.58-1353.33] [2604.62-2652.2] [4986.32-5061.59] 1029.26 2160.00 1355.47 2931.45 5379.83 35 2011 36 [1016.25-1042.40] [2140.46-2179.66] [1341.05-1370.02] [2906.32-2956.75] [5340.88-5419.00] 37 1348.85 1779.92 1483.27 2929.17 6028.88 2012 38 [1333.96-1363.86] [1762.16-1797.80] [1468.15-1498.50] [2903.97-2954.54] [5987.54-6070.44] 39 1466.16 2059.88 1723.64 4174.78 6535.45 2013 40 [1450.69-1481.75] [2040.76-2079.14] [1707.32-1740.09] [4144.47-4205.26] [6491.99-6579.14] 41 1520.52 2173.38 1843.66 5012.76 6174.78 42 2014 [1504.78-1536.37] [2153.71-2193.19] [1826.76-1860.68] [4979.45-5046.23] [6132.51-6217.26] 43 1397.14 2160.64 2339.49 5444.36 7303.19 44 2015 45 [1382.05-1412.36] [2141.00-2180.41] [2320.44-2358.66] [5409.71-5479.19] [7257.56-7349.03] 2145.78 2251.45 3072.88 6116.59 7186.53 46 2016 47 [2127.10-2164.59] [2231.38-2271.66] [3051.00-3094.88] [6079.75-6153.61] [7141.19-7232.08] 48 1948.73 3331.61 3278.79 7057.39 8964.66 2017 49 [1930.92-1966.66] [3307.18-3356.18] [3256.18-3301.51] [7017.81-7097.14] [8914.02-9015.51] 50 51 52 Table 4: Rate of underserved population by year and deprivation quintile 53 54 55 56 DISCUSSION 57 58 Primary care is considered to be the foundation of the health care system throughout 59 60 the world. Currently, in countries with rapidly ageing populations and a high number
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 23
1 2 3 of people living with chronic diseases, a strong and accessible base of primary care 4 5 providers is essential to the proper level of medical services as well to health 6 promotion for the whole population. There is a general agreement in the scientific 7 8 literature regarding a workforce crisis at the level of primary care; the shortage of 9 primary care physicians now calls for attention and urgency worldwide.64 10 11 In our present report, the following trends and phenomena are clearly identified as 12 13 characteristics of the Hungarian primary care workforce crisis in the decade examined: 14 15 - Increasing number of primary care practices with a vacant GP post 16 17 - Decreasing number of GPs, especially in mixed and paediatric practices 18 - Increasing averageFor agepeer of GPs inreview practice only 19 - Feminization of the profession indicated by increasing representation of female 20 21 GPs 22 - Strong association between deprived areas and practices with unfilled posts for 23 24 GP 25 - Intense movement of GPs from the deprived areas to the least-deprived ones 26 27 It is evident that the GP shortage has a direct effect on the quality of primary care 28 29 service provision, but it is reasonable to suppose that it causes problems at the level of 30 secondary care and in access to public health services. Although the usefulness and 31 32 effectiveness of the gatekeeping function of GPs in reducing utilization of emergency 33 and hospital services is a subject of intense discussion with contradictory 34 35 conclusions,see reviewed in 65-66 in the CEE countries with a very unfavourable 36 health status of their population,67-68 it seems to be essential. In these countries 37 38 (except the Czech Republic and Slovakia), patients do not have direct access to 39 hospital consultants, but the GPs control access to secondary care.69 Although, 40 41 currently, more outpatient specialist services can be approached in Hungary without 42 referral, the traditional referral system is still the dominant one in practice. 43 Consequently, the increasing number of practices with unfilled GP posts may strongly 44 45 affect the health status of the populations not only by missing primary care and 46 preventive services, but through a backlog of referrals to relevant specialists, as well. It 47 48 is worth mentioning that the population of the most deprived areas is the most affected 49 by the GP shortage and the replacement of GPs in practices especially in small rural 50 51 settlements is only partial, if any. 52 53 The chronic shortage of doctors in the primary health care sector owing to low pay, 54 restricted opportunities for training or career progression and poor working conditions 55 56 has been a well-known fact for a long time,70-71 but in the last decade, an 57 exponential increase in vacant GP posts was observed in different countries. The Pulse 58 59 magazine survey of 860 GPs in May 2017 found that 12.2% of positions were 60 currently vacant in England, an increase from 2.1% in 2011 when Pulse started
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 23 BMJ Open
1 2 3 collecting these data.72 Similarly, in a survey in Scotland, 22% of responding 4 5 primary care practices reported current GP vacancies as of 31 August 2015, an 6 increase from 9% of practices reporting current vacancies in 2013.73 The same 7 8 phenomenon can be detected in Hungary, where the number of primary care practices 9 with an unfilled GP post almost tripled in the period of 2010-2017. The decrease is 10 11 significant in the number of GPs in all types of practices, but it is more pronounced in 12 mixed and paediatric practices. It seems that the mixed general practices are less 13 14 attractive among GPs. One of the possible reasons of this phenomenon is that most of 15 these practices are found in small settlements where the GP is the only primary care 16 service provider. The reason of the decrease in the number of paediatricians is also 17 18 clear; the paediatricianFor medical peer specialization review training only focuses mostly on the hospital 19 replacement in Hungary. Hospitals in many parts of the country are currently reporting 20 21 shortages in paediatricians and young medical graduates take the hospital work more 22 attractive and the carrier paths more clear at present. 23 24 Parallel with the decrease in the number of GPs, the average age of the GPs became 25 26 significantly higher. The age distribution clearly indicates that the number of male GPs 27 under the age of 60 years was lower in 2017 than in 2007 across all ages, while in the 28 29 case of female GPs, although the trend is very similar, increases can be seen at certain 30 ages, mainly at younger ones. This fact is in harmony with the observed feminization 31 32 of the profession. In a systematic review based on 32 relevant publications to examine 33 the evidence that quantifies the effect of feminization on practice characteristics, 34 Hedden and her colleagues74 note “that female GPs self-report fewer hours of work 35 36 than male GPs, have fewer patient encounters, and deliver fewer services, but spend 37 longer with their patients during a contact and address more separate presenting 38 39 problems in one visit”, and conclude that “the feminization of the workforce is likely 40 to have a small negative impact on the availability of primary health care services, and 41 42 the drivers of observed differences between male and female GPs are complex and 43 nuanced.” Considering the fact that Hungary is among the 10 countries that have the 44 45 best parental leave policies in the world ― the length of paid maternity leave is 24 46 weeks and is paid at 70% of salary, which can be extended by an additional time off 47 paid at 40% of salary for 136 weeks75 ― female doctors may stay away from work 48 49 for a fairly long time, which can make finding a suitable temporary replacement 50 problematic. 51 52 53 Behind the fact that in Hungarian primary care there is a strong association between 54 deprived areas and practices with unfilled posts for GPs and the intense movement of 55 GPs from the deprived areas to the least-deprived ones is the probability of an existing 56 57 financial background. Although the majority of GPs are private entrepreneurs 58 capitation-based financing is almost exclusively based on the list size of the practices, 59 60 and only approximately 3% of GPs’ reimbursement is pay-for-performance.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 23
1 2 3 Consequently, GPs are financially motivated to maximize the number of persons in the 4 5 practice and minimize performance.39 According to the estimated data from the 6 Hungarian Medial Chamber, in 2017, after covering the expenses related to 7 8 maintaining the infrastructure and salaries of the other health staff (at least a nurse) the 9 remaining income that can be used to pay the salary of the GP was approximately 10 11 1000 EUR before taxation. Although it can be – and should be – refused on the basis 12 of ethical considerations, it remains a fact that the informal so called “gratuity” 13 payment from the patients for GP services is more or less essential to make a living. It 14 15 is reasonable to suppose that moving GPs from deprived areas to least-deprived ones is 16 ― at least in part ― motivated by the expected increase in “gratuity” payment. These 17 18 observations clearlyFor indicate peer that targeted review intervention only from the Ministry of Human 19 Capacities is needed to provide the National Institute of Health Insurance Fund 20 21 Management with decision-making responsibility and financing opportunities to 22 support effective recruitment and retention of GPs not only for the most-deprived 23 areas, but for the Hungarian primary care system as a whole, as well. 24 25 26 The recruitment and retention problems are well known in the primary care systems of 27 28 developed countries, but information about effective strategies and interventions is 29 very limited. Recently, a systematic review was undertaken in which 51 studies 30 31 assessing 42 interventions categorized into 13 groups (financial incentives, recruiting 32 rural students, international recruitment, rural- or primary-care-focused undergraduate 33 placements, rural or underserved postgraduate training, well-being or peer support 34 35 initiatives, marketing, mixed interventions, support for professional development or 36 research, retainer schemes, re-entry schemes, specialized recruiters or case managers 37 38 and delayed partnerships) were retrieved. Weak evidence supported the use of 39 postgraduate placements in underserved areas, undergraduate rural placements and 40 41 recruiting students to medical school from rural areas, and there was mixed evidence 42 about financial incentives.76 43 44 A single report on a longitudinal small area whole country study using lower super 45 46 output area data and the Index of Multiple Deprivation 2010 rankings for England 47 described substantial inequalities in GPs’ service, expressed in full-time equivalent 48 49 (FTE) GPs before the ‘Equitable Access to Primary Medical Care’ (EAPMC) 50 programme was launched in December 2007. EMPMC invested £250 million towards 51 52 establishing new general practices and GP-led health centres, as well as towards 53 extending opening hours and expanding services in the 38 most “underdoctored” 54 Primary Care Trust areas and providing a guaranteed competitive salaries for GPs. 55 56 Total numbers of FTE for GPs have grown much faster in the most-deprived fifth of 57 small areas in England than elsewhere, with the GP supply in the most affluent fifth 58 59 growing at the slowest pace over the 10 years covered by the study.16 Our findings 60 obtained in our longitudinal study strongly differ from the results presented as
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 23 BMJ Open
1 2 3 outcomes of the EAMC programme. It is reasonable to suppose that the main cause of 4 5 the difference that competitive salaries for GPs are not guaranteed in Hungary. 6 According to OECD figures, the income of a GP in Denmark is five times larger than 7 8 the income of a GP in Hungary, and the average GP in Scandinavia earns 5-10 times 9 more than the average GP in Hungary.77 In 2013, a salary increase programme for 10 11 GPs started in Hungary, but the annual growth rate in nominal terms was as low as 12 2.3%.78 However, since 2014, a single settlement allowance has been given to GPs 13 14 at the time of entering into a permanently vacant GP posts, no permanent incentives 15 were introduced to support movement by GPs into underserved deprived areas. 16 Although studies on the effect of financial incentives for GPs in Hungary are lacking, 17 18 it can be stated thatFor without peer increasing review the salaries of only GPs there is no chance to stop 19 the deepening workforce crisis in Hungarian primary care. 20 21 22 The limitation of our study is that it is an ecological one that assumes that individual 23 members of a group all have the average characteristics of the group as whole, when, 24 in fact, any association observed between variables at the group level does not 25 26 necessarily mean that the same association exists for an individual plucked from the 27 group. In addition, we had no opportunity to study the effects of different health care 28 29 reforms on the primary care workforce in Hungary. 30 31 32 FIGURE LEGENDS 33 34 35 Figure 1: The age and sex distribution of GPs in 2007 and 2017 in Hungary 36 37 Figure 2: The distribution pattern of deprivation (A) and of the relative vacancy rates 38 (B) in Hungary and the relationship between deprivation and the relative vacancy rate 39 (C) 40 41 42 43 44 REFERENCES 45 46 1 Campbell J, Dussault G, Buchan J, et al. A universal truth: no health without a 47 workforce. Forum report, Third Global Forum on Human Resources for Health, 48 Recife, Brazil. Geneva: World Health Organization, 2013. 49 50 https://www.who.int/workforcealliance/knowledge/resources/GHWA_AUniversal 51 TruthReport.pdf (accessed Feb 2019) 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 23
1 2 3 2 Hays R, Pong LT, Leopando et al. Primary care in the Asia-Pacific region: 4 5 challenges and solutions. Asia Pac Fam Med 2012;11:8. doi:10.1186/1447-056X- 6 11-8 7 3 Strasser R, Kam SM, Regalado SM. Rural health care access and policy in 8 developing countries. Annu Rev Public Health. 2016;37:395-412. doi: 9 10 10.1146/annurev-publhealth-032315-021507. Epub 2016 Jan 6. 11 4 Berwick DM, Hackbarth AD. Eliminating waste in US health care. JAMA. 12 2012;14:1513–6. doi: 10.1001/jama.2012.362. 13 5 Petterson, SM, Liaw, WR, Tran, C, et al. Estimating the Residency Expansion 14 15 Required to Avoid Projected Primary Care Physician Shortages by 2035. Ann Fam 16 Med 2015;2:107–14. doi:10.1370/afm.1760 17 18 6 Dale J, PotterFor R, Owen peer K, et al. review Retaining the generalonly practitioner workforce in 19 England: what matters to GPs? A cross-sectional study. BMC Fam Pract 20 21 2015;1:140. doi:10.1186/s12875-015-0363-1. 22 7 Hayhoe B, Majeed A, Hamlyn M, et al. Primary care workforce crisis: how 23 many more GPs do we need? RCGP Annual Conference, Harrogate, 2016. 24 8 Majeed A. Shortage of general practitioners in the NHS. BMJ 2017;358:j3191. 25 26 doi: https://doi.org/10.1136/bmj.j3191 27 9 Burla L, Widmer M. Le corps médical en Suisse: effectif et évolutions jusqu’en 28 2011. Neuchâtel: Observatoire suisse de la santé; 2012. 29 https://www.obsan.admin.ch/sites/default/files/publications/2015/obsan_bulletin_2 30 31 012-03_f.pdf (accessed Feb 2019) 32 10 Seematter-Bagnoud L, Junod J, Jaccard Ruedin H, et al. Offre et recours aux 33 soins médicaux ambulatoires en Suisse – Projections à l’horizon 2030. Neuchâtel: 34 Observatoire suisse de la santé; 2008. 35 36 https://www.obsan.admin.ch/sites/default/files/publications/2015/arbeitsdokument- 37 33.pdf (accessed Feb 2019) 38 11 OECD/EU (2016), Health at a Glance: Europe 2016: State of Health in the EU 39 Cycle. OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264265592-en 40 41 12 Comino EJ, Davies GP, Krastev Y, et al. A systematic review of interventions 42 to enhance access to best practice primary health care for chronic disease 43 management, prevention and episodic care. BMC Health Serv Res 2012;12:415 doi: 44 10.1186/1472-6963-12-415 45 46 13 IMO Position Paper on Health Inequalities, Irish Medical Association. 2012. 47 https://www.imo.ie/policy-international-affair/overview/IMO-Position-Paper-on- 48 Health-Inequalities.pdf (accessed Feb 2019) 49 50 14 Bauer J, Brueggmann D, Ohlendorf D, et al. General practitioners in German 51 metropolitan areas - distribution patterns and their relationship with area level 52 measures of the socioeconomic status. BMC Health Serv Res. 2016 Nov 53 25;16(1):672. doi: 10.1186/s12913-016-1921-5 54 55 15 Baker M, Ware J, Morgan K. Time to put patients first by investing in general 56 practice. Br J Gen Pract 2014;64:268–9. doi:10.3399/bjgp14X679921 57 16 Asaria M, Cookson R, Fleetcroft R, et al. Unequal socioeconomic distribution 58 of the primary care workforce: whole-population small area longitudinal study. 59 60 BMJ Open 2016;1:e008783. doi: 10.1136/bmjopen-2015-008783
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 23 BMJ Open
1 2 3 17 Steinhaeuser J, Otto P, Goetz K, et al. Rural area in a European country from a 4 5 health care point of view: an adaption of the Rural Ranking Scale. BMC Health 6 Serv Res. 2014;14:147. doi: 10.1186/1472-6963-14-147. 7 18 Kringos D, Boerma W, Hutchinson A, et al. Building primary care in a 8 changing Europe. 38th ed. Copenhagen; 2015 9 10 http://www.euro.who.int/__data/assets/pdf_file/0018/271170/BuildingPrimaryC 11 areChangingEurope.pdf (accessed Feb 2019) 12 19 Yardim M, Uner S. Geographical disparities in the distribution of physicians in 13 Turkey. TAF Prev Med Bull 2013;12:487–94. doi: 10.5455/pmb.1-1351692762 14 15 20 Matsumoto M, Inoue K, Farmer J, et al. Geographic distribution of primary care 16 physicians in Japan and Britain. Health Place 2010;16:164–6. doi: 17 10.1016/j.healthplace.2009.07.005 18 21 Petterson SM,For Phillips peer RL Jr, Bazemore review AW, et only al. Unequal distribution of the 19 20 US primary care workforce. Am Fam Physician 2013;87;11, pp. 21 http://www.aafp.org/afp/2013/0601/od1.pdf (accessed Feb 2019) 22 22 Hann M, Gravelle H. The maldistribution of general practitioners in England 23 and Wales: 1974–2003. Br Gen Pract 2004;54:894–8. 24 25 23 Isabel C, Paula V. Geographic distribution of physicians in Portugal. Eur J 26 Health Econ 2010;11:383–93. doi: 10.1007/s10198-009-0208-8 27 24 Fülöp G, Kopetsch T, Hofstätter G, et al. Regional distribution effects of ‘needs 28 planning’ for office-based physicians in Germany and Austria-methods and 29 30 empirical findings. J Public Health 2008;16:447–55. doi: 10.1007/s10389-008- 31 0187-8 32 25 Brown M. Using Gini-style indices to evaluate the spatial patterns of health 33 practitioners: Theoretical considerations and an application based on Alberta data. 34 35 Soc Sci Med 1994;38:1243–56. doi: 10.1016/0277-9536(94)90189-9 36 26 Stapleton G, Schröder-Bäck P, Brand H, et al. Health inequalities and regional 37 specific scarcity in primary care physicians: ethical issues and criteria. Int J Public 38 Health 2014;59:449–55. doi: 10.1007/s00038-013-0497-7 39 40 27 World Health Organization, The World Health Report - Primary health care 41 (now more than ever). World Health Organization, 2008. 42 https://www.who.int/whr/2008/en/ (accessed Feb 2019) 43 28 World Health Organization, Increasing access to health workers in remote and 44 45 rural areas through improved retention. Background paper. Geneva, 2009. 46 https://www.who.int/hrh/retention/guidelines/en/ (accessed Feb 2019) 47 29 Watson J, Humphrey A, Peters Klimm F, et al. Motivation and satisfaction in 48 49 GP training: a UK cross-sectional survey. Br J Gen Pract 2011; 591: e645–9. doi: 50 10.3399/bjgp11X601352 51 30 Fletcher E, Abel GA, Anderson R, et al. Quitting patient care and career break 52 intentions among general practitioners in South West England: findings of a census 53 54 survey of general practitioners. BMJ Open 2017, doi: 10.1136/bmjopen-2017- 55 015853 56 31 Blane DN, McLean G, Watt G. Distribution of GPs in Scotland by age, gender 57 and deprivation. Scott Med J 2015;4:214-9. doi:10.1177/0036933015606592. Epub 58 2015 Sep 23. 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 23
1 2 3 32 Iacobucci G. GPs in Northern Ireland face "full blown crisis," BMA warns. 4 5 BMJ. 2016 Jun 7;353:i3202. doi: 10.1136/bmj.i3202 6 33 Zieler P. Physician shortage in Saxony-Anhalt. The former youngster returns as 7 country doctor. MMW Fortschr Med. 2015;17:36-7. 8 34 Schmidt S, Gresser U. Development and consequences of physician shortages 9 10 in Bavaria. Versicherungsmedizin. 2014;1:25-9. 11 35 Roeger LS, Reed RL, Smith BP. Equity of access in the spatial distribution of 12 GPs within an Australian metropolitan city. Aust J Prim Health. 2010;4:284-90. 13 doi: 10.1071/PY10021. 14 15 36 Ádány R, Kósa K, Sándor J, et al. General practitioners' cluster: a model to reorient 16 primary health care to public health services. Eur J Public Health. 2013; 23:529-530. 17 doi: 10.1093/eurpub/ckt095. 18 37 Sándor J, KósaFor K, Fürjes peer G, et al. review Public health services only provided in the framework 19 20 of general practitioners' clusters. Eur J Public Health. 2013; 23:530-532. doi: 21 10.1093/eurpub/ckt096. 22 38 Kósa K, Sándor J, Dobos É, et al. Human resources development for the operation 23 of general practitioners' cluster. Eur J Public Health. 2013; 23:532-533. doi: 24 25 10.1093/eurpub/ckt097. 26 39 Sándor J, Kósa K, Papp M, et al. Capitation-Based Financing Hampers the 27 Provision of Preventive Services in Primary Health Care. Front Public Health 28 2016;4:200, doi: 10.3389/fpubh.2016.00200. eCollection 2016 29 30 40 Lember M, Cartier T, Bourgueil Y. Structure and organization of primary care. 31 In: Kringos DS, Boerma WGW, Hutchinson A (eds), Building primary care in a 32 changing Europe, Observatory Studies Series, No. 38. 2015, p. 41-9. 33 41 Juhasz A, Nagy C, Paldy A. Development of a deprivation index and its relation 34 35 to premature mortality due to diseases of the circulatory system in Hungary, 1998- 36 2004. Soc Sci Med 2010;70:1342–9. doi: 10.1016/j.socscimed.2010.01.024. 37 42 Nagy C, Juhasz A, Beale L, et al. Mortality amenable to health care and its 38 relation to socio-economic status in Hungary, 2004-08. Eur J Public Health 39 2012;22:620–4. doi: 10.1093/eurpub/ckr143 40 41 43 Nagy C, Juhasz A, Papp Z, et al. Hierarchical spatio-temporal mapping of 42 premature mortality due to alcoholic liver disease in Hungary, 2005-2010. Eur J 43 Public Health 2014;24:827–833. doi: 10.1093/eurpub/ckt169 44 44 Boruzs K, Juhasz A, Nagy C, et al. Relationship between statin utilization and 45 46 socioeconomic deprivation in Hungary. Front Pharmacol 2016;7:66. doi: 47 10.3389/fphar.2016.00066 48 45 Jakab Zs, Juhasz A, Nagy C, et al. Trends and territorial inequalities of 49 incidence and survival of childhood leukaemia and their relations to socioeconomic 50 51 status in Hungary, 1971–2015. Eur J Cancer Prev 1. 2017. doi: 52 10.1097/CEJ.0000000000000386. 53 46 Fairburn J, Maier W, Braubach M. Incorporating Environmental Justice into 54 Second Generation Indices of Multiple Deprivation: Lessons from the UK and 55 56 Progress Internationally. Int J Environ Res Public Health 2016;13:750. 57 doi:10.3390/ijerph13080750 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 23 BMJ Open
1 2 3 47 Department for Communities and Local Government: English Indices of 4 5 Deprivation 2015. http://www.gov.uk/government/statistics/english-indices-of- 6 deprivation-2015 (accessed Feb 2019) 7 48 The Scottish Government Scottish Index of Multiple Deprivation. 8 (accessed Feb 2019) 9 http://www.gov.scot/Topics/Statistics/SIMD 10 49 Welsh Government Welsh Index of Multiple Deprivation (WIMD). 11 http://gov.wales/statistics-and-research/welsh-index-multiple- 12 deprivation/?lang=en (accessed Feb 2019) 13 14 50 Northern Ireland Multiple Deprivation Measure 2017 (NIMDM2017). 15 https://www.nisra.gov.uk/statistics/deprivation/northern-ireland-multiple- 16 deprivation-measure-2017-nimdm2017 (accessed Feb 2019) 17 51 Havard S, Deguen S, Bodin J, et al. A small-area index of socioeconomic 18 For peer review only 19 deprivation to capture health inequalities in France. Soc Sci Med 2008;67:2007–16. 20 doi:10.1016/j.socscimed.2008.09.031 21 52 Sánchez-Cantalejo C, Ocana-Riola R, Fernández-Ajuria A. Deprivation index 22 23 for small areas in Spain. Soc Indic Res 2008;89:259–273. doi: 10.1007/s11205- 24 007-9114-6 25 53 Domínguez-Berjón MF, Borrell C, Cano-Serral G, et al. Constructing a 26 deprivation index based on census data in large Spanish cities(the MEDEA 27 28 project). Gac Sanit 2008;22:179–87. doi:https://doi.org/10.1157/13123961 29 54 Kroll LE, Schumann M, Hoebel J, et al. Regional health differences – 30 developing a socioeconomic deprivation index for Germany. Journal of Health 31 Monitoring Published Online First: 2017. doi: 10.17886/RKI-GBE-2017-048 32 55 Meijer M, Engholm G, Grittner U, et al. A socioeconomic deprivation index for 33 34 small areas in Denmark. Scand J Public Health 2013;41:560–9. 35 doi:10.1177/1403494813483937 36 56 Panczak R, Galobardes B, Voorpostel M, et al. A Swiss neighbourhood index 37 of socioeconomic position: development and association with mortality. J 38 39 Epidemiol Community Health 2012;66:1129. doi:10.1136/jech-2011-200699 40 57 Caranci N, Biggeri A, Grisotto L, et al. L’indice di deprivazione italiano a 41 livello di sezione di censimento: definizione, descrizione e associazione con la 42 mortalità. Epidemiol Prev 2010;34:167–76. 43 44 58 Šlachtová H, Tomášková H, Šplíchalová A, et al. Czech socio-economic 45 deprivation index and its correlation with mortality data. International Journal of 46 Public Health 2009;54:267–273. 47 59 Zadnik V, Guillaume E, Lokar K, et al. Slovenian Version of The European 48 49 Deprivation Index at Municipal Level. Zdr Varst 2018;57:47–54. doi:10.2478/sjph- 50 2018-0007 51 60 Besag J, York J, and Mollié A, A Bayesian image restoration with two 52 applications in spatial statistics. Ann Inst Stat Math.1991;43:1–20. doi: 53 54 10.1007/BF00116466 55 61 R Development Core Team. R: A Language and Environment for Statistical 56 Computing, R Foundation for Statistical Computing, Vienna, Austria, 2017. 57 www.R-project.org (accessed Feb 2019) 58 59 62 Rue H, Martino S, Chopin N. Approximate Bayesian inference for latent 60 Gaussian models using integrated nested Laplace approximations (with discussion).
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 23
1 2 3 Journal of the Royal Statistical Society, Series B, 2009;2:319-392, www.r-inla.org 4 5 (accessed Feb 2019) 6 63 Beale L, Hodgson S, Abellan JJ, et al. Evaluation of spatial relationships 7 between health and the environment: the rapid inquiry facility. Environ Health 8 Perspect 2010;118, 1306–12. doi: 10.1289/ehp.0901849 9 10 64 Song Z, Chopra V, McMahon LF. Addressing the Primary Care Workforce 11 Crisis. Am J Manag Care 2015;8:e452-4. 12 65 Van den Heede K, Van de Voorde C. Interventions to reduce emergency 13 department utilisation: A review of reviews. Health Policy 2016;120:1337-1349. 14 15 doi: 10.1016/j.healthpol.2016.10.002. 16 66 Greenfield G, Foley K, Majeed A. Rethinking primary care's gatekeeper role. 17 BMJ 2016;i4803. doi:10.1136/bmj.i4803 18 67 KaranikolosFor M, Adany peer R, McKee review M. The epidemiological only transition in Eastern 19 20 and Western Europe: a historic natural experiment. Eur J Public Health 21 2017;suppl_4:4-8. doi: 10.1093/eurpub/ckx158 22 68 Mackenbach JP. Nordic paradox, Southern miracle, Eastern disaster: 23 persistence of inequalities in mortality in Europe. Eur J Public Health 24 25 2017;suppl_4:14-17. doi: 10.1093/eurpub/ckx160 26 69 Euro Health Consumer Index 2016, Health Consumer Powerhouse Ltd, 2017. 27 https://healthpowerhouse.com/media/EHCI-2016/EHCI-2016-report.pdf (accessed 28 Feb 2019) 29 30 70 Rechel B, Dubois C-A, McKee M (eds). The Health Care Workforce in Europe 31 Learning from experience. European Observatory on Health Systems and Policies, 32 2006. http://www.euro.who.int/__data/assets/pdf_file/0008/91475/E89156.pdf 33 (accessed Feb 2019) 34 35 71 Kuehn BM. Reports Warn of Primary Care Shortages. JAMA. 2008;300:1872– 36 5. doi: 10.1001/jama.300.16.1872. 37 72 NHSProvider. Summary of vacancy, shortfall and fill rate data for the clinical 38 workforce. November 2017. https://nhsproviders.org/media/3903/summary-of- 39 40 vacancy-shortfall-and-fill-rate-data-for-the-clinical-workforce.pdf (accessed Feb 41 2019) 42 73 Primary Care Workforce Survey Scotland 2015, A Survey of Scottish General 43 Practices and General Practice Out of Hours Services. National Services Scotland, 44 45 Information Services Division, Publication Report, 14 June, 2016 46 http://www.isdscotland.org/Health-Topics/General-Practice/Publications/2016-06- 47 14/2016-06-14-PrimaryCareWorkforceSurveyScotland2015-Report.pdf (accessed 48 Feb 2019) 49 50 74 Heeden L, Barer ML, Cardiff K, et al. The implications of the feminization of 51 the primary care physician workforce on service supply: a systematic review. 52 Human Resources for Health 2014;12:32. https://doi.org/10.1186/1478-4491-12-32 53 75 Weller C.These 10 countries have the best parental leave policies in the world, 54 55 World Economic Forum, 2016. 56 https://www.weforum.org/agenda/2016/08/these-10-countries-have-the-best- 57 parental-leave-policies-in-the-world (accessed 28 Feb 2019) 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 23 BMJ Open
1 2 3 76 Verma P, Ford JA, Stuart A, et al. A systematic review of strategies to recruit 4 5 and retain primary care doctors. BMC Health Serv Res 2016;16:126. 6 https://doi.org/10.1186/s12913-016-1370-1 7 77 OECD Health Statistics 2017. 8 78 OECD Health Statistics 2015. 9 10 11 12 13 14 FOOTNOTES 15 16 Contributors MP, JS, CN, AJ and RA developed the study design. LK identified the 17 18 practices by location,For vacancy peer and characteristics review of GPs. only MP, AJ and CN did the data 19 processing and analyses. MP wrote the first draft of the paper and RA finalized it. All 20 authors critically revised the first draft and approved the final manuscript submitted for 21 publication. 22 23 Funding This work was supported by the Swiss Contribution SH/8/1 project and by 24 the GINOP-2.3.2-15-2016-00005 project financed by the European Union under the 25 European Social Fund and European Regional Development Fund, as well as by the 26 27 Hungarian Academy of Sciences (TK2016-78). 28 29 Competing interests None declared. 30 31 Patient consent Not required. 32 33 Ethics approval Not required. 34 35 Provenance and peer review Not commissioned; externally peer reviewed. 36 37 Data sharing statement Data about general practices (number and type of practices, 38 distribution, vacancy, served population) and general practitioners (number, age, sex, 39 distribution) for the time period from 1 January 2007 to 1 January 2017 were provided 40 by the National Institute of Health Insurance Fund Management. Our data about the 41 42 population number at the municipality level were obtained from the Regional 43 Informational System of the Ministry of Local Government and Regional 44 Development, while data regarding the the level of education, the rate of 45 unemployment, the rate of one-parent families, the rate of large families, the density of 46 47 housing and car ownership were obtained from the Hungarian Central Statistical 48 Office and the Hungarian Tax and Financial Control Administration. These data are 49 available from the above institutions upon request. 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 170x175mm (300 x 300 DPI) 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 23 BMJ Open
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 170x250mm (300 x 300 DPI) 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml