Gustav Kjellsson
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Gustav Kjellsson has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Lina Maria Ellegård
Department of Economics, Lund University
7 shared publications - Ulf-Goran Gerdtham
Lund University
4 shared publications - Dennis Petrie
Centre for Health Economics, Monash University, Melbourne, Australia
3 shared publications - Beata Borgström Bolmsjö
Center for Primary Health Care Research, Department of Clinical Sciences, Malmö, Lund University
2 shared publications - Jens Dietrichson
VIVE – The Danish Center for Social Science Research
2 shared publications - Moa Wolff
Center for Primary Health Care Research, Department of Clinical Sciences Malmö, Lund University, Malmö, Sweden
2 shared publications
Latest Publications
- Open Access
Prospective evaluation of a self-report-guided strategy for targeted coronary artery calcium imaging
Background: Coronary artery calcium (CAC) imaging directly assesses subclinical calcified coronary atherosclerosis, but population-wide imaging is not recommended. Simple pre-screening may help identify individuals most likely to benefit from CAC imaging. We previously developed a self-report-based model to estimate the probability of CAC ≥100. This study prospectively evaluated a strategy based on this model to select individuals for CAC imaging. We assessed agreement between model-predicted probability and observed prevalence of CAC ≥100 among participants undergoing computed tomography (CT) imaging and examined patterns of preventive lipid-lowering therapy. Methods: The PRedict and Identify cOronary atherosclerosis–Now (PRIO–Now) study applied a prospective, two-step, population-based screening approach. Individuals aged 59–60 years were invited to complete a self-report questionnaire. Eligible respondents without previous ischemic heart disease whose model-predicted probability of CAC ≥100 exceeded the predefined threshold were invited to clinical assessment and non-contrast coronary CT imaging. The primary analysis assessed agreement between model-predicted probabilities and the observed prevalence of CAC ≥100 among CT completers. Results: Of 8,000 invited individuals, 2,588 (32%) completed the questionnaire. Of 2,375 eligible respondents, 814 were classified as high risk and 563 underwent CT imaging. Among CT completers, the mean predicted probability of CAC ≥100 was 28.3% (95% CI 27.2–29.3), compared with an observed prevalence of 28.4% (95% CI 24.8–32.4), corresponding to an expected/observed ratio of 0.99 and a Brier score of 0.19. Among participants with CAC ≥100, 64% were not receiving lipid-lowering therapy and 11% had LDL-C ≥1.8 mmol/L. Conclusions: A self-report-guided strategy enabled targeted CAC imaging in a model-selected cohort. Among participants completing CT imaging, the observed prevalence of CAC ≥100 was comparable with the mean model-predicted probability. These findings suggest that self-report data may support pre-selection for CAC imaging and help identify opportunities for preventive treatment among individuals with elevated CAC.
Open AccessProspective evaluation of a self-report-guided strategy for targeted coronary artery calcium imaging
Background: Coronary artery calcium (CAC) imaging directly assesses subclinical calcified coronary atherosclerosis, but population-wide imaging is not recommended. Simple pre-screening may help identify individuals most likely to benefit from CAC imaging. We previously developed a self-report-based model to estimate the probability of CAC ≥100. This study prospectively evaluated a strategy based on this model to select individuals for CAC imaging. We assessed agreement between model-predicted probability and observed prevalence of CAC ≥100 among participants undergoing computed tomography (CT) imaging and examined patterns of preventive lipid-lowering therapy. Methods: The PRedict and Identify cOronary atherosclerosis–Now (PRIO–Now) study applied a prospective, two-step, population-based screening approach. Individuals aged 59–60 years were invited to complete a self-report questionnaire. Eligible respondents without previous ischemic heart disease whose model-predicted probability of CAC ≥100 exceeded the predefined threshold were invited to clinical assessment and non-contrast coronary CT imaging. The primary analysis assessed agreement between model-predicted probabilities and the observed prevalence of CAC ≥100 among CT completers. Results: Of 8,000 invited individuals, 2,588 (32%) completed the questionnaire. Of 2,375 eligible respondents, 814 were classified as high risk and 563 underwent CT imaging. Among CT completers, the mean predicted probability of CAC ≥100 was 28.3% (95% CI 27.2–29.3), compared with an observed prevalence of 28.4% (95% CI 24.8–32.4), corresponding to an expected/observed ratio of 0.99 and a Brier score of 0.19. Among participants with CAC ≥100, 64% were not receiving lipid-lowering therapy and 11% had LDL-C ≥1.8 mmol/L. Conclusions: A self-report-guided strategy enabled targeted CAC imaging in a model-selected cohort. Among participants completing CT imaging, the observed prevalence of CAC ≥100 was comparable with the mean model-predicted probability. These findings suggest that self-report data may support pre-selection for CAC imaging and help identify opportunities for preventive treatment among individuals with elevated CAC.
- Open Access
- 23 June 2026
- Published by Taylor & Francis in Scandinavian Journal of Primary Health Care
- Vol. 44 (1) , 2690608
Continuity in a fragmented healthcare system: – organizational and individual determinants
Continuity of care is key for high-quality primary care, associated with improved health outcomes, reduced mortality and more efficient use of resources. Achieving relational continuity is challenging in systems with high provider choice and fragmented care-seeking. Sweden exemplifies this tension, combining broad access to primary care and low GP continuity. To examine how individual characteristics (age, morbidity, socioeconomic status and migration background) and primary care center (PCC) features (ownership, size, physician turnover, patient mix and location) are associated with relational continuity and to contrast within-PCC continuity with total continuity systemwide. Retrospective cohort study using linked administrative register data covering all in-person physician contacts in primary care for 1.4 million residents in Region Skåne, Sweden. Continuity of Care Index (CoCI) was measured using all primary care physician visits over 36-months. Linear regression models estimated associations between individual and PCC characteristics and continuity, adjusting for individual- and PCC-level covariates. The characteristics most strongly associated with higher continuity were low physician turnover, older age, chronic conditions, smaller PCC size and private ownership. Individuals with higher education, higher income and foreign background had lower total continuity. Differences in continuity were more strongly associated with PCC characteristics than with patient characteristics. Relational continuity in Swedish primary care is associated primarily with organizational factors, particularly physician turnover and practice size. Fragmented care-seeking among specific groups, especially individuals born outside the Nordic countries, contributes to lower total continuity but does not reflect weaker patient-provider relationships within PCCs.
- 23 June 2026
- Published by Taylor & Francis in Scandinavian Journal of Primary Health Care
- Vol. 44 (1) , 2690608
Open AccessContinuity in a fragmented healthcare system: – organizational and individual determinants
Continuity of care is key for high-quality primary care, associated with improved health outcomes, reduced mortality and more efficient use of resources. Achieving relational continuity is challenging in systems with high provider choice and fragmented care-seeking. Sweden exemplifies this tension, combining broad access to primary care and low GP continuity. To examine how individual characteristics (age, morbidity, socioeconomic status and migration background) and primary care center (PCC) features (ownership, size, physician turnover, patient mix and location) are associated with relational continuity and to contrast within-PCC continuity with total continuity systemwide. Retrospective cohort study using linked administrative register data covering all in-person physician contacts in primary care for 1.4 million residents in Region Skåne, Sweden. Continuity of Care Index (CoCI) was measured using all primary care physician visits over 36-months. Linear regression models estimated associations between individual and PCC characteristics and continuity, adjusting for individual- and PCC-level covariates. The characteristics most strongly associated with higher continuity were low physician turnover, older age, chronic conditions, smaller PCC size and private ownership. Individuals with higher education, higher income and foreign background had lower total continuity. Differences in continuity were more strongly associated with PCC characteristics than with patient characteristics. Relational continuity in Swedish primary care is associated primarily with organizational factors, particularly physician turnover and practice size. Fragmented care-seeking among specific groups, especially individuals born outside the Nordic countries, contributes to lower total continuity but does not reflect weaker patient-provider relationships within PCCs.
- Open Access
- 11 February 2026
- Published by Elsevier in SSRN Electronic Journal
Well-informed choices? Effects of information interventions in primary care on care quality
Market frictions, such as imperfect information or hassle costs, may reduce the efficiency of consumer choice in healthcare settings. We use data from two randomised policy interventions in a Swedish region, which improved the access to information about primary care providers and reduced the switching costs of one percent of the adult population and of a sample of new residents. We examine the intervention effects on a large number of clinical process quality measures, access to care, and adverse health events, measured at the individual level. We find no significant effect of the interventions on any of the outcomes.
- 11 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessWell-informed choices? Effects of information interventions in primary care on care quality
Market frictions, such as imperfect information or hassle costs, may reduce the efficiency of consumer choice in healthcare settings. We use data from two randomised policy interventions in a Swedish region, which improved the access to information about primary care providers and reduced the switching costs of one percent of the adult population and of a sample of new residents. We examine the intervention effects on a large number of clinical process quality measures, access to care, and adverse health events, measured at the individual level. We find no significant effect of the interventions on any of the outcomes.
- Open Access
- 19 November 2025
- Published by Taylor & Francis in Scandinavian Journal of Primary Health Care
- Vol. 44 (1) , 1-10
GP turnover in a multiprofessional team-based primary care system: evidence from Sweden
GP recruitment and retention difficulties challenge the traditional general practice model. Task-shifting and relieving GPs from financial risk have been suggested to make primary care more attractive. In Sweden's multiprofessional team-based primary care system, GPs usually work as salaried employees and there is extensive task-shifting. Salaried employment facilitates mobility, potentially leading to high turnover. The opportunity to work on fixed contracts can also increase turnover rates. To describe practice turnover rates and examine associations with practice characteristics in a Swedish region. Analysis of observational register data from Skåne, Sweden (1.4 million residents). Turnover rates were calculated for 157 primary care practices in 2010-2018. The main dataset included all physicians - permanent and temporary workers - regularly providing care in each month. To understand the role of temporary workers, a supplementary analysis was performed on permanently employed GPs and registrars at 80 public practices in 2019. Associations between turnover and practice characteristics were examined in bivariate analyses and multiple regressions. Annual practice turnover rates ranged between 20-40% (mean 30%), showing no time trend. The high rates mainly reflected the use of temporary GPs; in the supplementary analysis of permanent GPs and registrars, the mean annual turnover rate in 2019 was 13-15%. Turnover was higher for practices with socially deprived patients or high workload. Private practices had lower turnover conditional on the higher workload. The results indicate that a primary care system with salaried GPs facilitates GP mobility, which in turn creates barriers to continuity of care.
- 19 November 2025
- Published by Taylor & Francis in Scandinavian Journal of Primary Health Care
- Vol. 44 (1) , 1-10
Open AccessGP turnover in a multiprofessional team-based primary care system: evidence from Sweden
GP recruitment and retention difficulties challenge the traditional general practice model. Task-shifting and relieving GPs from financial risk have been suggested to make primary care more attractive. In Sweden's multiprofessional team-based primary care system, GPs usually work as salaried employees and there is extensive task-shifting. Salaried employment facilitates mobility, potentially leading to high turnover. The opportunity to work on fixed contracts can also increase turnover rates. To describe practice turnover rates and examine associations with practice characteristics in a Swedish region. Analysis of observational register data from Skåne, Sweden (1.4 million residents). Turnover rates were calculated for 157 primary care practices in 2010-2018. The main dataset included all physicians - permanent and temporary workers - regularly providing care in each month. To understand the role of temporary workers, a supplementary analysis was performed on permanently employed GPs and registrars at 80 public practices in 2019. Associations between turnover and practice characteristics were examined in bivariate analyses and multiple regressions. Annual practice turnover rates ranged between 20-40% (mean 30%), showing no time trend. The high rates mainly reflected the use of temporary GPs; in the supplementary analysis of permanent GPs and registrars, the mean annual turnover rate in 2019 was 13-15%. Turnover was higher for practices with socially deprived patients or high workload. Private practices had lower turnover conditional on the higher workload. The results indicate that a primary care system with salaried GPs facilitates GP mobility, which in turn creates barriers to continuity of care.
- Open Access
- 15 July 2025
- Published by Elsevier in Social Science & Medicine
- Vol. 383, 118351
Patient choice and socioeconomic disparities in the quality of healthcare: Evidence from Swedish registry data
- 15 July 2025
- Published by Elsevier in Social Science & Medicine
- Vol. 383, 118351
Open AccessPatient choice and socioeconomic disparities in the quality of healthcare: Evidence from Swedish registry data
- Open Access
- 2 February 2022
- Published by JMIR Publications Inc. in JMIR Human Factors
- Vol. 9 (1) , e33034
Trends in Remote Health Care Consumption in Sweden: Comparison Before and During the First Wave of the COVID-19 Pandemic
Background: Remote assessment of respiratory tract infections (RTIs) has been a controversial topic during the fast development of private telemedicine providers in Swedish primary health care. The possibility to unburden the traditional care has been put against a questionable quality of care as well as risks of increased utilization and costs. The COVID-19 pandemic has contributed to a changed management of patient care to decrease viral spread, with an expected shift in contact types from in-person to remote ones. Objective: The main aim of this study was to compare health care consumption and type of contacts (in-person or remote) for RTIs before and during the COVID-19 pandemic. The second aim was to study whether the number of follow-up contacts after an index contact for RTIs changed during the study period, and whether the number of follow-up contacts differed if the index contact was in-person or remote. A third aim was to study whether the pattern of follow-up contacts differed depending on whether the index contact was with a traditional or a private telemedicine provider. Methods: The study design was an observational retrospective analysis with a description of all index contacts and follow-up contacts with physicians in primary care and emergency rooms in a Swedish region (Skåne) for RTIs including patients of all ages and comparison for the same periods in 2018, 2019, and 2020. Results: Compared with 2018 and 2019, there were fewer index contacts for RTIs per 1000 inhabitants in 2020. By contrast, the number of follow-up contacts, both per 1000 inhabitants and per index contact, was higher in 2020. The composition of both index and follow-up contacts changed as the share of remote contacts, in particular for traditional care providers, increased. Conclusions: During the COVID-19 pandemic in 2020, fewer index contacts for RTIs but more follow-up contacts were conducted, compared with 2018-2019. The share of both index and follow-up contacts that were conducted remotely increased. Further studies are needed to study the reasons behind the increase in remote contacts, and if it will last after the pandemic, and more clinical guidelines for remote assessments of RTI are warranted.
- 2 February 2022
- Published by JMIR Publications Inc. in JMIR Human Factors
- Vol. 9 (1) , e33034
Open AccessTrends in Remote Health Care Consumption in Sweden: Comparison Before and During the First Wave of the COVID-19 Pandemic
Background: Remote assessment of respiratory tract infections (RTIs) has been a controversial topic during the fast development of private telemedicine providers in Swedish primary health care. The possibility to unburden the traditional care has been put against a questionable quality of care as well as risks of increased utilization and costs. The COVID-19 pandemic has contributed to a changed management of patient care to decrease viral spread, with an expected shift in contact types from in-person to remote ones. Objective: The main aim of this study was to compare health care consumption and type of contacts (in-person or remote) for RTIs before and during the COVID-19 pandemic. The second aim was to study whether the number of follow-up contacts after an index contact for RTIs changed during the study period, and whether the number of follow-up contacts differed if the index contact was in-person or remote. A third aim was to study whether the pattern of follow-up contacts differed depending on whether the index contact was with a traditional or a private telemedicine provider. Methods: The study design was an observational retrospective analysis with a description of all index contacts and follow-up contacts with physicians in primary care and emergency rooms in a Swedish region (Skåne) for RTIs including patients of all ages and comparison for the same periods in 2018, 2019, and 2020. Results: Compared with 2018 and 2019, there were fewer index contacts for RTIs per 1000 inhabitants in 2020. By contrast, the number of follow-up contacts, both per 1000 inhabitants and per index contact, was higher in 2020. The composition of both index and follow-up contacts changed as the share of remote contacts, in particular for traditional care providers, increased. Conclusions: During the COVID-19 pandemic in 2020, fewer index contacts for RTIs but more follow-up contacts were conducted, compared with 2018-2019. The share of both index and follow-up contacts that were conducted remotely increased. Further studies are needed to study the reasons behind the increase in remote contacts, and if it will last after the pandemic, and more clinical guidelines for remote assessments of RTI are warranted.
- Open Access
- 23 August 2021
- Published by JMIR Publications Inc. in JMIR Preprints
Trends in Remote Health Care Consumption in Sweden: Comparison Before and During the First Wave of the COVID-19 Pandemic (Preprint)
BACKGROUND Remote assessment of respiratory tract infections (RTIs) has been a controversial topic during the fast development of private telemedicine providers in Swedish primary health care. The possibility to unburden the traditional care has been put against a questionable quality of care as well as risks of increased utilization and costs. The COVID-19 pandemic has contributed to a changed management of patient care to decrease viral spread, with an expected shift in contact types from in-person to remote ones. OBJECTIVE The main aim of this study was to compare health care consumption and type of contacts (in-person or remote) for RTIs before and during the COVID-19 pandemic. The second aim was to study whether the number of follow-up contacts after an index contact for RTIs changed during the study period, and whether the number of follow-up contacts differed if the index contact was in-person or remote. A third aim was to study whether the pattern of follow-up contacts differed depending on whether the index contact was with a traditional or a private telemedicine provider. METHODS The study design was an observational retrospective analysis with a description of all index contacts and follow-up contacts with physicians in primary care and emergency rooms in a Swedish region (Skåne) for RTIs including patients of all ages and comparison for the same periods in 2018, 2019, and 2020. RESULTS Compared with 2018 and 2019, there were fewer index contacts for RTIs per 1000 inhabitants in 2020. By contrast, the number of follow-up contacts, both per 1000 inhabitants and per index contact, was higher in 2020. The composition of both index and follow-up contacts changed as the share of remote contacts, in particular for traditional care providers, increased. CONCLUSIONS During the COVID-19 pandemic in 2020, fewer index contacts for RTIs but more follow-up contacts were conducted, compared with 2018-2019. The share of both index and follow-up contacts that were conducted remotely increased. Further studies are needed to study the reasons behind the increase in remote contacts, and if it will last after the pandemic, and more clinical guidelines for remote assessments of RTI are warranted.
- 23 August 2021
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessTrends in Remote Health Care Consumption in Sweden: Comparison Before and During the First Wave of the COVID-19 Pandemic (Preprint)
BACKGROUND Remote assessment of respiratory tract infections (RTIs) has been a controversial topic during the fast development of private telemedicine providers in Swedish primary health care. The possibility to unburden the traditional care has been put against a questionable quality of care as well as risks of increased utilization and costs. The COVID-19 pandemic has contributed to a changed management of patient care to decrease viral spread, with an expected shift in contact types from in-person to remote ones. OBJECTIVE The main aim of this study was to compare health care consumption and type of contacts (in-person or remote) for RTIs before and during the COVID-19 pandemic. The second aim was to study whether the number of follow-up contacts after an index contact for RTIs changed during the study period, and whether the number of follow-up contacts differed if the index contact was in-person or remote. A third aim was to study whether the pattern of follow-up contacts differed depending on whether the index contact was with a traditional or a private telemedicine provider. METHODS The study design was an observational retrospective analysis with a description of all index contacts and follow-up contacts with physicians in primary care and emergency rooms in a Swedish region (Skåne) for RTIs including patients of all ages and comparison for the same periods in 2018, 2019, and 2020. RESULTS Compared with 2018 and 2019, there were fewer index contacts for RTIs per 1000 inhabitants in 2020. By contrast, the number of follow-up contacts, both per 1000 inhabitants and per index contact, was higher in 2020. The composition of both index and follow-up contacts changed as the share of remote contacts, in particular for traditional care providers, increased. CONCLUSIONS During the COVID-19 pandemic in 2020, fewer index contacts for RTIs but more follow-up contacts were conducted, compared with 2018-2019. The share of both index and follow-up contacts that were conducted remotely increased. Further studies are needed to study the reasons behind the increase in remote contacts, and if it will last after the pandemic, and more clinical guidelines for remote assessments of RTI are warranted.
- Open Access
- 18 March 2020
- Published by Wiley in Health Economics
- Vol. 29 (6) , 716-730
Patient choice, entry, and the quality of primary care: Evidence from Swedish reforms
Policies aiming to spur quality competition among health care providers are ubiquitous, but their impact on quality is ex ante ambiguous, and credible empirical evidence is lacking in many contexts. This study contributes to the sparse literature on competition and primary care quality by examining recent competition enhancing reforms in Sweden. The reforms aimed to stimulate patient choice and entry of private providers across the country but affected markets differently depending on the initial market structure. We exploit the heterogeneous impact of the reforms in a difference‐in‐differences strategy, contrasting more and less exposed markets over the period 2005–2013. Although the reforms led to substantially more entry of new providers in more exposed markets, the effects on primary care quality were modest: We find small improvements of patients' overall satisfaction with care, but no consistently significant effects on avoidable hospitalisation rates or satisfaction with access to care. We find no evidence of economically meaningful quality reductions on any outcome measure.
- 18 March 2020
- Published by Wiley in Health Economics
- Vol. 29 (6) , 716-730
Open AccessPatient choice, entry, and the quality of primary care: Evidence from Swedish reforms
Policies aiming to spur quality competition among health care providers are ubiquitous, but their impact on quality is ex ante ambiguous, and credible empirical evidence is lacking in many contexts. This study contributes to the sparse literature on competition and primary care quality by examining recent competition enhancing reforms in Sweden. The reforms aimed to stimulate patient choice and entry of private providers across the country but affected markets differently depending on the initial market structure. We exploit the heterogeneous impact of the reforms in a difference‐in‐differences strategy, contrasting more and less exposed markets over the period 2005–2013. Although the reforms led to substantially more entry of new providers in more exposed markets, the effects on primary care quality were modest: We find small improvements of patients' overall satisfaction with care, but no consistently significant effects on avoidable hospitalisation rates or satisfaction with access to care. We find no evidence of economically meaningful quality reductions on any outcome measure.
- 10 July 2017
- Published by Wiley in Health Economics
- Vol. 27 (2) , 440-447
Extending decomposition analysis to account for unobserved heterogeneity and persistence in health behavior: Income‐related smoking inequality among Swedish women
This article suggests an enrichment of the standard method for decomposition of the concentration index to account for unobserved heterogeneity and persistence in health behavior. As the underlying regression model in the decomposition, this approach uses a dynamic random‐effect probit that both consider individual heterogeneity, using a Mundlak type of specification, and applies a simple solution to account for smoking persistence. I illustrate the suggested approach using a panel of Swedish women in Statistics Sweden's Survey of Living Conditions for one vital health‐related behavior, smoking. The results highlight the importance of persistence and show that education and living in a single household are the main drivers of income‐related smoking inequality.
- 10 July 2017
- Published by Wiley in Health Economics
- Vol. 27 (2) , 440-447
Extending decomposition analysis to account for unobserved heterogeneity and persistence in health behavior: Income‐related smoking inequality among Swedish women
This article suggests an enrichment of the standard method for decomposition of the concentration index to account for unobserved heterogeneity and persistence in health behavior. As the underlying regression model in the decomposition, this approach uses a dynamic random‐effect probit that both consider individual heterogeneity, using a Mundlak type of specification, and applies a simple solution to account for smoking persistence. I illustrate the suggested approach using a panel of Swedish women in Statistics Sweden's Survey of Living Conditions for one vital health‐related behavior, smoking. The results highlight the importance of persistence and show that education and living in a single household are the main drivers of income‐related smoking inequality.
- 1 July 2017
- Published by Wolters Kluwer Health in Epidemiology
- Vol. 28 (4) , 604-607
Graphical Tools for Monitoring Inequality
- 1 July 2017
- Published by Wolters Kluwer Health in Epidemiology
- Vol. 28 (4) , 604-607
Graphical Tools for Monitoring Inequality