Dataset Information

A framework for exploring non-response patterns over time in health surveys.

ABSTRACT:

Background

Most health surveys have experienced a decline in response rates. A structured approach to evaluate whether a decreasing - and potentially more selective - response over time biased estimated trends in health behaviours is lacking. We developed a framework to explore the role of differential non-response over time. This framework was applied to a repeated cross-sectional survey in which the response rate gradually declined.

Methods

We used data from a survey conducted biannually between 1995 and 2017 in the city of Rotterdam, The Netherlands. Information on the sociodemographic determinants of age, sex, and ethnicity was available for respondents and non-respondents. The main outcome measures of prevalence of sport participation and watching TV were only available for respondents. The framework consisted of four steps: 1) investigating the sociodemographic determinants of responding to the survey and the difference in response over time between sociodemographic groups; 2) estimating variation in health behaviour over time; 3) comparing weighted and unweighted prevalence estimates of health behaviour over time; and 4) comparing associations between sociodemographic determinants and health behaviour over time.

Results

The overall response rate per survey declined from 47% in 1995 to 15% in 2017. The probability of responding was higher among older people, females, and those with a Western background. The response rate declined in all subgroups, and a faster decline was observed among younger persons and those with a non-Western ethnicity as compared to older persons and those with a Western ethnicity. Variation in health behaviours remained constant. Prevalence estimates and associations did not follow the changes in response over time. On the contrary, the difference in probability of participating in sport gradually decreased between males and females, while no differential change in the response rate was observed.

Conclusions

Providing insights on non-response patterns over time is essential to understand whether declines in response rates may have influenced estimated trends in health behaviours. The framework outlined in this study can be used for this purpose. In our example, in spite of a major decline in response rate, there was no evidence that the risk of non-response bias increased over time.

SUBMITTER: Molenberg FJM

PROVIDER: S-EPMC7890886 | biostudies-literature | 2021 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A framework for exploring non-response patterns over time in health surveys.

Mölenberg Famke J M FJM de Vries Chris C Burdorf Alex A van Lenthe Frank J FJ

BMC medical research methodology 20210218 1

<h4>Background</h4>Most health surveys have experienced a decline in response rates. A structured approach to evaluate whether a decreasing - and potentially more selective - response over time biased estimated trends in health behaviours is lacking. We developed a framework to explore the role of differential non-response over time. This framework was applied to a repeated cross-sectional survey in which the response rate gradually declined.<h4>Methods</h4>We used data from a survey conducted b ...[more]

PMID: 33602123

Similar Datasets

Project description:BackgroundLittle is known about how individuals engage with electronic health (eHealth) interventions over time and whether this engagement predicts health outcomes.ObjectiveThe objectives of this study, by using the example of a specific type of eHealth intervention (ie, websites for smoking cessation), were to determine (1) distinct groups of log-in trajectories over a 12-month period, (2) their association with smoking cessation, and (3) baseline user characteristics that predict trajectory group membership.MethodsWe conducted a functional clustering analysis of 365 consecutive days of log-in data from both arms of a large (N=2637) randomized trial of 2 website interventions for smoking cessation (WebQuit and Smokefree), with a primary outcome of 30-day point prevalence smoking abstinence at 12 months. We conducted analyses for each website separately.ResultsA total of 3 distinct trajectory groups emerged for each website. For WebQuit, participants were clustered into 3 groups: 1-week users (682/1240, 55.00% of the sample), 5-week users (399/1240, 32.18%), and 52-week users (159/1240, 12.82%). Compared with the 1-week users, the 5- and 52-week users had 57% higher odds (odds ratio [OR] 1.57, 95% CI 1.13-2.17; P=.007) and 124% higher odds (OR 2.24, 95% CI 1.45-3.43; P<.001), respectively, of being abstinent at 12 months. Smokefree users were clustered into 3 groups: 1-week users (645/1309, 49.27% of the sample), 4-week users (395/1309, 30.18%), and 5-week users (269/1309, 20.55%). Compared with the 1-week users, 5-week users (but not 4-week users; P=.99) had 48% higher odds (OR 1.48, 95% CI 1.05-2.07; P=.02) of being abstinent at 12 months. In general, the WebQuit intervention had a greater number of weekly log-ins within each of the 3 trajectory groups as compared with those of the Smokefree intervention. Baseline characteristics associated with trajectory group membership varied between websites.ConclusionsPatterns of 1-, 4-, and 5-week usage of websites may be common for how people engage in eHealth interventions. The 5-week usage of either website, and 52-week usage only of WebQuit, predicted a higher odds of quitting smoking. Strategies to increase eHealth intervention engagement for 4 more weeks (ie, from 1 week to 5 weeks) could be highly cost effective.Trial registrationClinicalTrials.gov NCT01812278; https://www.clinicaltrials.gov/ct2/show/NCT01812278 (Archived by WebCite at http://www.webcitation.org/6yPO2OIKR).

Project description:IntroductionHIV testing is the entry point to HIV prevention, care and treatment and needs continuous evaluation to understand whether all social groups have accessed services equally. Addressing disparities in HIV testing between social groups results in effective and efficient response against HIV prevention. Despite these benefits, there was no previous study on inequality and determinants over time in Ethiopia. Thus, the objective of this research was to examine socioeconomic inequality in individuals undertaking HIV testing over time, allowing for the identification of persistent and emerging determinants.MethodsData sources for the current study were the 2011 and 2016 Ethiopian Demographic Health Surveys. The 2016 population health survey is the one that Ethiopia used to set national AIDS response strategies; there was no other recent survey with HIV/AIDS-related indicators in Ethiopia. The final sample size for the current study was 28,478 for the year 2011 and 25,542 for the year 2016. The concentration curve and Erreygers' concentration index were used to estimate socioeconomic inequality in HIV testing. Subsequently, decomposition analysis was performed to identify persistent and emerging contributors of socioeconomic inequality. Generalized linear regression model with the logit link function was employed to estimate the marginal effect, elasticity, Erreygers' concentration index (ECI), and absolute and percentage contributions of each covariate.ResultsThe concentration curve was below the line of equality over time, revealing the pro-rich inequality in HIV testing. The inequality was observed in both 2011 (ECI = 0.200) and 2016 (ECI = 0.213). A household wealth rank had the highest percentage contribution (49.2%) for inequality in HIV testing in 2011, which increased to 61.1% in 2016. Additional markers include listening to the radio (13.4% in 2011 and 12.1% in 2016), education status (8.1% in 2011 and 6.8% in 2016), and resident (-2.0% in 2011 and 6.3% in 2016). Persistent determinants of individuals undertaking HIV testing were age 20-34 years, geographic region, education status, marital status, religion, income, media exposure (listening to the radio, reading newspaper, watching television), knowledge about HIV/AIDS, and attitudes towards people living with HIV. Age between 35 and 44 years and urban residence emerged as new associated factors in 2016.ConclusionsThe higher HIV testing coverage was among individuals with higher socioeconomic status in Ethiopia. Socioeconomic inequality amongst individuals undertaking HIV testing was diverging over time. Household wealth rank, mass media exposure, education status, and resident took the largest share in explaining the disparity in individuals undertaking HIV testing between the lower and higher income groups. Therefore, interventions to equalise HIV testing coverage should take account of these determinants.

Project description:ObjectivesTo describe trends in caesarean sections and facilities performing caesareans over time in Tanzania and examine the readiness of such facilities in terms of infrastructure, equipment and staffing.DesignNationally representative, repeated cross-sectional surveys of women and health facilities.SettingTanzania.ParticipantsWomen of reproductive age and health facility staff.Main outcome measuresPopulation-based caesarean rate, absolute annual number of caesareans, percentage of facilities reporting to perform caesareans and three readiness indicators for safe caesarean care: availability of consistent electricity, 24 hour schedule for caesarean and anaesthesia providers, and availability of all general anaesthesia equipment.ResultsThe caesarean rate in Tanzania increased threefold from 2% in 1996 to 6% in 2015-16, while the total number of births increased by 60%. As a result, the absolute number of caesareans increased almost fivefold to 120 000 caesareans per year. The main mechanism sustaining the increase in caesareans was the doubling of median caesarean volume among public hospitals, from 17 caesareans per month in 2006 to 35 in 2014-15. The number of facilities performing caesareans increased only modestly over the same period. Less than half (43%) of caesareans in Tanzania in 2014-15 were performed in facilities meeting the three readiness indicators. Consistent electricity was widely available, and 24 hour schedules for caesarean and (less systematically) anaesthesia providers were observed in most facilities; however, the availability of all general anaesthesia equipment was the least commonly reported indicator, present in only 44% of all facilities (34% of public hospitals).ConclusionsGiven the rising trend in numbers of caesareans, urgent improvements in the availability of general anaesthesia equipment and trained anaesthesia staff should be made to ensure the safety of caesareans. Initial efforts should focus on improving anaesthesia provision in public and faith-based organisation hospitals, which together perform more than 90% of all caesareans in Tanzania.

Project description:INTRODUCTION:Demographic and Health Surveys, widely used for estimation of fertility and reproductive health indicators in developing countries, remain underutilized for the study of pregnancy termination. This is partly due to most surveys not reporting the type of pregnancy termination, whether spontaneous or induced. Reproductive calendar data makes it possible to examine termination patterns according to contraceptive use at the time of pregnancy. Contraceptive failure is expected to increase the likelihood of induced abortion helping in the interpretation of reported termination patterns. MATERIALS AND METHODS:We use individual-level calendar data regarding 623,966 pregnancies to analyze levels and differentials in reported patterns of pregnancy termination by age, union status, and contraceptive use in 107 DHS surveys from 50 countries. From the estimates of the probability of pregnancy termination, we compute derived reproductive health indicators providing an assessment of what is driving the differences by comparison to the few surveys reporting the type of pregnancy termination. RESULTS:From our estimates, 10.9% of pregnancies do not end in live-birth and 63.7% of them are spontaneous terminations. Reported pregnancy termination is higher among women using contraceptives, consistent with expectations. Very low levels of reported PT in some countries, particularly in sub-Saharan Africa, suggests possible underreporting. Differential patterns emerging from cluster analysis and regional rates indicate high rates of pregnancy termination driven by induced abortion in countries from the Former Soviet Union and Asian countries with liberal laws. Most countries with restrictive abortion laws have low levels of reported termination. While the probabilities of pregnancy termination are higher at older ages, termination rates generally peak at younger ages due to higher conception rates. DISCUSSION:This is the first large comparative study of the patterns of reported pregnancy termination in DHS surveys. While we have explored the extent to which differences arise from spontaneous terminations or induced abortion, more research is needed regarding the determinants of reported pregnancy termination.

Dataset Information

A framework for exploring non-response patterns over time in health surveys.

Background

Methods

Results

Conclusions

Publications

A framework for exploring non-response patterns over time in health surveys.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets