Dataset Information

Can the validity of a cohort be improved by reweighting based on register data? Evidence from the Swedish MDC study

ABSTRACT:

Background

In any study with voluntary participation, self-selection risks leading to invalid conclusions. If the determinants of selection are observed, it is however possible to restore the parameters of interest by reweighting the sample to match the population, but this approach has seldom been applied in epidemiological research.

Methods

We reweighted the Malmö Diet and Cancer (MDC) study based on population register data on background variables, including socio-demographics and hospital admissions for both participants and the background population. Following individuals from baseline in 1991–1996 and at most until 2016, we studied mortality (all-cause, cancer, and CVD), incidences (cancer and CVD), and associations between these outcomes and background variables. Results from the unweighted and reweighted participant sample were compared with those from the background population.

Results

Mortality was substantially lower in participants than in the background population, but reweighting the sample helped only little to make the numbers similar to those in the background population. For incidences and associations, numbers were generally similar between participants and the background population already without reweighting, rendering reweighting unnecessary.

Conclusion

Reweighting samples based on an extensive range of sociodemographic characteristics and previous hospitalizations does not necessarily yield results that are valid for the population as a whole. In the case of MDC, there appear to be important factors related to both mortality and selection into the study that are not observable in registry data, making it difficult to obtain accurate numbers on population mortality based on cohort participants. These issues seem less relevant for incidences and associations, however. Overall, our results suggest that representativeness must be judged on a case-by-case basis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-020-10004-z.

SUBMITTER: Nilsson A

PROVIDER: S-EPMC7747383 | biostudies-literature | 2020 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:IntroductionThe Swedish Register of Palliative Care (SRPC) is a national quality register that collects data about end-of-life care from healthcare providers that care for dying patients. Data are used for quality control and research. Data are mainly collected with an end-of-life questionnaire (ELQ), which is completed by healthcare staff after the death of a patient. A previous validity assessment of the ELQ showed insufficient validity in some items including symptom relief. The aim of this study was to examine the validity of the revised ELQ.Materials and methodsData from 100 consecutive patients' medical records at two specialised palliative care units were used to complete new ELQs, which were then compared to the ELQ registrations from the SRPC for the same patients. The level of agreement was calculated for each ELQ item. To account for the possibility of the agreement occurring by chance, Cohen's kappa was calculated for suitable items. To examine the extent of registration mistakes when transferring the paper form to the web, the original paper versions of the ELQ filled out at the units were compared to data from the ELQs reported to the SRPC.ResultsLevel of agreement between ELQ registrations from the SRPC and the new ELQs based on the medical records varied between 0.55 and 1.00, where 24 items showed level of agreement above 0.80 and 9 items showed level of agreement below 0.80. Cohen's kappa with 95% confidence intervals was calculated for 24 items. The kappa values showed that two items had poor agreement, four fair agreement, 11 moderate agreement, five good agreement and two very good agreement. The level of agreement varied between 0.93 and 1.00 when comparing the ELQ registrations in the SRPC and the paper forms.ConclusionThe revised ELQ contains more items with high levels of agreement between registrations in the SRPC and notes in the patients' medical records when compared to the previous version. Validating issues around symptom assessment remains a challenge in our model of quality assessment.

Project description:BackgroundIn cohorts with voluntary participation, participants may not be representative of the underlying population, leading to distorted estimates. If the relevant sources of selective participation are observed, it is however possible to restore the representativeness by reweighting the sample to resemble the target population. So far, few studies in epidemiology have applied reweighting based on extensive register data on socio-demographics and disease history, or with self-reported data on health and health-related behaviors.MethodsWe examined selective participation at baseline and the first two follow-ups of the Scania Public Health Cohort (SPHC), a survey conducted in Southern Sweden in 1999/2000 (baseline survey; n = 13,581 participants, 58% participation rate), 2005 (first follow-up, n = 10,471), and 2010 (second follow-up; n = 9,026). Survey participants were reweighted to resemble the underlying population with respect to a broad range of socio-demographic, disease, and health-related characteristics, and we assessed how selective participation impacted the validity of associations between self-reported overall health and dimensions of socio-demographics and health.ResultsParticipants in the baseline and follow-up surveys were healthier and more likely to be female, born in Sweden, middle-aged, and have higher socioeconomic status. However, the differences were not very large. In turn, reweighting the samples to match the target population had generally small or moderate impacts on associations. Most examined regression coefficients changed by less than 20%, with virtually no changes in the directions of the effects.ConclusionOverall, selective participation with respect to the observed factors was not strong enough to substantially alter the associations with self-assessed health. These results are consistent with an interpretation that SPHC has high validity, perhaps reflective of a relatively high participation rate. Since validity must be determined on a case-by-case basis, however, researchers should apply the same method to other health cohorts to assess and potentially improve the validity.

Project description:ObjectiveTo investigate if unemployment during an economic downturn is associated with mortality, even among men with markers of better health (higher cognitive function scores and qualifications), and to assess whether the associations vary by age at unemployment.DesignLongitudinal register-based cohort study.SettingStudy entry was in 1990 and 2001 when Sweden was entering periods of significant economic contraction.ParticipantsA representative sample of men from the general population (n=234 782) born between 1952 and 1956 who participated in military conscription examinations. Men in receipt of disability or sickness benefit at study entry were excluded.Main outcome measureAll-cause mortality.ResultsUnemployment compared with employment in 1991 (ages 34-38 years) produced adjusted HRs (with 95% CIs) for all-cause mortality (3651 deaths) during follow-up to 2001 and after stratification by education of 2.35 (1.99 to 2.76) for compulsory education, 2.25 (1.97 to 2.58) for up to 3 years postcompulsory education and 1.90 (1.40 to 2.57) for more than 3 years postcompulsory education. When unemployment was compared with employment in 2001 (ages 45-49 years) with follow-up to 2010, the pattern of mortality risk (4271 deaths) stratified by education was reversed, producing adjusted HRs of 2.81 (2.47 to 3.21) for compulsory education, 2.87 (2.58 to 3.19) for up to 3 years postcompulsory education and 3.44 (2.78 to 4.25) for more than 3 years postcompulsory education. Interaction testing confirmed effect modification by age/period (p=0.003). The degree of gradient reversal was slightly less pronounced after stratification by cognitive function but produced a similar pattern of results (p=0.004).ConclusionsUnemployment at older ages is associated with greater mortality risk than at younger ages, with the greatest relative increase in risk among men with markers of better health, suggesting the greater vulnerability of all older workers to unemployment-associated exposures.

Project description:OBJECTIVES:Studies have found a 'healthy-migrant effect' (HME) among arriving migrants, that is, a better health status compared with others in the home country, but also in comparison with the population in the host country. The aims were to investigate whether the HME hypothesis is applicable to the Swedish context, that is, if health outcomes differed between a group of mainly labour migrants (Western migrants) and a group of mainly refugee/family reunion migrants (non-Western migrants) compared with the native Swedish population, and if there were any correlations between labour market attachment (LMA) and these health outcomes. DESIGN:Register-based, longitudinal cohort study. PARTICIPANTS:The cohort was defined on 31 December 1990 and consisted of all migrants aged 18-47 years who arrived in Sweden in 1985-1990 (n=74 954) and a reference population of native Swedes (n=1 405 047) in the same age span. They were followed for three consecutive 6-year periods (1991-1996, 1997-2002 and 2003-2008) and were assessed for five measures of health: hospitalisation for cardiovascular and psychiatric disorders, mortality, disability pension, and sick leave. RESULTS:Western migrants had, compared with native Swedes, lower or equal HRs for all health measures during all time periods, while non-Western migrants displayed higher or equal HRs for all health measures, except for mortality, during all time periods. Age, educational level, occupation and LMA explained part of the difference between migrants and native Swedes. High LMA was associated with higher HRs for cardiovascular disorders among Western migrants, higher HRs of psychiatric disorders among non-Western migrants and higher HRs of mortality among both migrant groups compared with native Swedes. CONCLUSIONS:There were indications of a HME among Western migrants, while less proof of a HME among non-Western migrants. Stratification for LMA and different migrant categories showed some interesting differences, and measurements of the HME may be inconclusive if not stratified by migrant category or other relevant variables.