Article Text
Abstract
Purpose To study 21st century trends in healthcare utilisation by patients with rheumatoid arthritis (RA) compared with the general population.
Methods Observational cohort study. Using Swedish healthcare register data, we identified 3977 Region Skåne residents (mean age in 2001, 62.7 years; 73% women) presenting with RA (International Classification of Diseases-10 codes M05 or M06) in 1998–2001. We randomly sampled two referents from the general population per RA patient matched for age, sex and area of residence. We calculated the year 2001–2010 trends for the annual ratio (RA cohort/referents) of the mean number of hospitalisations and outpatient clinic visits.
Results By the end of the 10-year period, 62% of patients and 74% of referents were still alive and resident in the region. From 2001 to 2010, the ratio (RA cohort/referents) of the mean number of hospitalisations for men and women decreased by 27% (p=0.01) and 28% (p=0.004), respectively. The corresponding decrease was 29% (p=0.005) and 16% (p=0.004) for outpatient physician care, 34% (p=0.009) and 18% (p=0.01) for nurse visits, and 34% (p=0.01) and 28% (p=0.004) for physiotherapy. The absolute reduction in number of hospitalisations was from an annual mean of 0.79 to 0.69 in male patients and from 0.71 to 0.59 in female patients. The corresponding annual mean number of consultations in outpatient physician care by male and female RA patients changed from 9.2 to 7.7 and from 9.9 to 8.7, respectively.
Conclusions During the first decade of the 21st century, coinciding with increasing use of earlier and more active RA treatment including biological treatment, overall inpatient and outpatient healthcare utilisation by a cohort of patients with RA decreased relative to the general population.
- Rheumatoid Arthritis
- Health services research
- Epidemiology
Statistics from Altmetric.com
Introduction
Since late 20th century, there have been a number of new effective pharmacological treatment options for patients with rheumatoid arthritis (RA). This has led to better control of inflammation, less structural joint damage, and reduced symptoms. Thus, earlier diagnosis, earlier and more active pharmacological treatment,1 ‘treat to target’ strategies,2 and the implementation of new treatment guidelines,3 as well as the use of biological agents early in the disease course, has changed the treatment paradigm for RA over the first decade of the 21st century.1 This may potentially have changed the need for healthcare in patients with RA unrelated to other major changes in the healthcare system affecting all patients.
In Sweden, 77–92% of patients receiving disease-modifying antirheumatic drugs use methotrexate.4 On a national level, ∼20% are receiving biological agents.5 Patients with early RA treated according to new regimens could be expected to have less impairment of function and work ability6 ,7 and fewer hospitalisations.8 ,9 However, screening and close monitoring of patients is also necessary when applying the new treatment regimens,10 which may impact on healthcare utilisation. Although there are several added benefits from the new treatments,11–13 it is difficult to predict what the future will bring in the long term in RA-related healthcare, since evidence on actual healthcare utilisation is scarce and to some extent conflicting.14
Our objective, using comprehensive observational register data, was to analyse healthcare utilisation with a focus on rheumatology and orthopaedics over the last decade in a closed cohort of patients with RA compared with a reference cohort from the general population.
Methods
The Skåne Healthcare Register
In Sweden, healthcare is based on a tax-financed system and is free of charge (except for a minor co-payment). All healthcare providers, both public and private, are required to submit information for reimbursement purposes. In the Region Skåne, the southernmost county of Sweden, with one-eighth of the Swedish population, all inpatient and outpatient visits are registered in the Skåne Healthcare Register (SCHR) by the patient's personal identifier.15 For all healthcare providers, date of visit and information on healthcare provider is recorded. For public care, diagnostic codes are registered according to the International Classification of Diseases (ICD) 10 system.
The national population register
Vital events (date of birth and death, marriage and change of residential address) of all inhabitants of Sweden are registered in the population register by the personal identification number.
Cohort definitions
RA cohort
Using SCHR data for the period 1998 to 2001, we identified all adult (≥18 years) female or male residents of the Region Skåne who had received a diagnosis of seropositive or other RA (ICD-10 codes M05 and/or M06) on at least two separate occasions, at least one of which was by a specialist in rheumatology or internal medicine (or under specialty training in those disciplines).
Reference cohort
Using the population register at the end of the year 2000, we randomly sampled two referents from the general population per RA patient matched by birth year, sex and area of residence.
Survival and healthcare utilisation
Using the population register, we traced residence status and survival for each subject (in both the RA cohort and the reference cohort) in the period 2001–2010, and using the SCHR we studied healthcare utilisation by each individual including hospitalisations and outpatient visits to physicians, nurses and physiotherapists (both in general practice and specialised care). When a subject died or relocated out of the county, data were censored from that time.
Statistical analysis
We analysed the annual mean number of hospitalisations in total and at a clinic of rheumatology/internal medicine or orthopaedics during the 10-year time frame. We also analysed the mean number of outpatient clinic visits to physicians (including general practitioners), nurses and physiotherapists for each calendar year. To evaluate possible trends relative to the background population, we calculated the ratio of the mean number of visits between the RA cohort and reference cohort for each calendar year and performed a test for trend across ordered groups. We considered a two-tailed p value of 0.05 or less to be significant (Stata software V.11.2).
The study was approved by the regional ethics committee of Lund University.
Results
General description of cohorts
At the beginning of the study period (1 January 2001), the RA cohort consisted of 3977 subjects (mean age 62.7 years, 72.7% women). By the end of the study period (31 December 2010), 2471 subjects (62.1%) were still alive and resident in the Region Skåne. During follow-up, 1417 (35.6%) subjects had died, 13 of whom had died after they had relocated, and 89 (2.2%) (still alive) had relocated out of the county (see online appendix A).
Of the 7954 matched referents, 1810 (22.8%) had died by end of the study period and 237 (3.0%) (still alive) had relocated from the Region Skåne. Hence, by the end of the study period, 5907 subjects (74.3%) remained alive and resident in the county (see online appendix B).
Hospitalisations
The annual mean number of hospitalisations in the RA cohort tended not to increase or decrease during follow-up (figure 1). However, there was a statistically significant trend for a reduced hospitalisation ratio between patients with RA and their reference subjects, both in men (−27%, p=0.01) and women (−28%, p=0.004) (table 1), suggesting a relative decrease in the number of hospitalisations among patients with RA compared with the general population.
The ratios for hospitalisations at a clinic of rheumatology or internal medicine displayed no trend in men (−16%, p=0.13), but declined in women (−51%, p=0.004) (table 1, figure 2). The hospitalisation ratio to an orthopaedic clinic decreased in both male (−39%, p=0.006) and female (−36%, p=0.02) patients (table 1, figure 2).
Outpatient healthcare
Visits to a physician
During follow-up, the annual mean number of outpatient visits to a physician decreased by approximately one visit in both male and female patients with RA (figure 1). The ratios between patients with RA and reference subjects also decreased for both men (−29%, p=0.005) and women (−16%, p=0.003) (table 1).
The outpatient consultation ratio to a rheumatologist and/or specialist in internal medicine suggested no significant trends (table 1, figure 2). The consultation ratio to a specialist in orthopaedics tended to decrease in women (−19%, p=0.04) (table 1).
Visits to a nurse
The mean number of visits to nurses (primary care and specialised care) increased for both male and female patients with RA, as well as for reference subjects (figure 1). However, the consultation ratio between both male (−34%, p=0.009) and female (−18%, p=0.01) patients with RA and their reference subjects decreased during follow-up (table 1).
Visits to a physiotherapist
The total number of consultations with physiotherapists (primary care and specialised care) decreased from an annual mean of 5.5 per female patient in 2001 to 3.4 in 2010 (figure 1). The pattern in male patients was similar. The physiotherapy consultation ratios between patients with RA and reference subjects decreased during follow-up in both men (−34%, p=0.01) and women (−28%, p=0.004, table 1).
Discussion
Major changes in the treatment of RA have occurred during the first decade of the 21st century. Using Swedish observational data, we found both inpatient and outpatient healthcare utilisation by a cohort of patients with RA to have decreased relative to the general population.
Increased need for healthcare is to be expected in an ageing group of people whether they have a chronic disease or not.16 We observed such tendencies in our reference subjects, but not to the same degree in the RA cohort. The decrease in the gap in healthcare utilisation between the RA cohort and referents seems largely to be explained by fewer visits by patients with RA to physicians outside the field of rheumatology, where we also saw fewer visits to physiotherapists and nurses. Less utilisation of physician care in the USA has previously been reported,17 and patients monitored closely in anti-TNF studies have also been reported to have less frequent visits to physicians.13 However, the explanation for these findings was traced to administration of anti-TNF therapy, and visits to physicians increased slightly for patients receiving therapy subcutaneously.13 ,14 In our study setting, monitoring of eventual biological treatment is integrated into the specialised outpatient care by both rheumatologists and nurses. Previous studies on secular trends on utilisation of nurses and physiotherapists are lacking, although this kind of information is important to understand and monitor the burden of RA.
Earlier studies on healthcare utilisation in RA have mainly been based on investigation of cost of disease, analysis of the cost of pharmacological treatment, or burden of disease. Importantly, observational studies based on existing data from the healthcare system have been encouraged to add value and new perspectives.14 ,18 Our results are based on actual observed healthcare utilisation and placed in context by comparison with reference subjects from the general population. However, there are also important limitations. In our cohort, we followed ∼80% of the total expected number of patients with RA resident in the Region Skåne.5 The remaining 20% of patients with RA, not captured, were predominantly patients who were regularly seen by private practising rheumatologists or did not consult a rheumatologist during 1998–2001, and their disease activity was probably less severe.19 ,20 Further, there is a healthy survivor effect—that is, patients with RA who survive the study period are, in general, healthier than those who do not. However, we can expect the same phenomenon in the reference cohort. Ultimately, we would have preferred to study a dynamic RA cohort compared with a dynamic reference cohort. However, increased reporting of diagnostic codes to the register during the study period, primarily from primary care, precluded such a study design, as it would introduce selection bias (probably capturing more and less severe RA cases, which may have affected the results). Finally, it is plausible that changes in the healthcare system (eg, more ambulatory treatment of certain conditions) during this decade may have affected the healthcare utilisation by patients with chronic and non-chronic diseases differently.
To conclude, during the 21st century (representing modern rheumatology), we found reduced overall healthcare utilisation in a cohort of patients with RA compared with the general population.
Acknowledgments
We would like to acknowledge statistical support from Ljuba Kedza, RC Syd, Sweden.
References
Supplementary materials
Lay summary
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online appendix 1
- Data supplement 2 - Online appendix 2
Footnotes
-
Contributors All authors participated in planning and performing this study. All authors confirm that they have read and approved the content presented in the final manuscript for submission.
-
Funding This work was supported by the Norrbacka-Eugenia Foundation, the Capio's Research Foundation, the Swedish Research Council, the Region Skåne and the Faculty of Medicine, Lund University, Sweden.
-
Competing interests None.
-
Ethics approval The regional ethics committee of Lund University.
-
Provenance and peer review Not commissioned; externally peer reviewed.