Involving hard-to-reach ethnic minorities in low-budget health research: lessons from a health survey among Moluccans in the Netherlands

Evaluation of a Wastewater-Based Epidemiological Approach to Estimate the Prevalence of SARS-CoV-2 Infections and the Detection of Viral Variants in Disparate Oregon Communities at City and Neighborhood Scales

BACKGROUND

Positive correlations have been reported between wastewater SARS-CoV-2 concentrations and a community's burden of infection, disease or both. However, previous studies mostly compared wastewater to clinical case counts or nonrepresentative convenience samples, limiting their quantitative potential.

OBJECTIVES

This study examined whether wastewater SARS-CoV-2 concentrations could provide better estimations for SARS-CoV-2 community prevalence than reported cases of COVID-19. In addition, this study tested whether wastewater-based epidemiology methods could identify neighborhood-level COVID-19 hotspots and SARS-CoV-2 variants.

METHODS

Community SARS-CoV-2 prevalence was estimated from eight randomized door-to-door nasal swab sampling events in six Oregon communities of disparate size, location, and demography over a 10-month period. Simultaneously, wastewater SARS-CoV-2 concentrations were quantified at each community's wastewater treatment plant and from 22 Newport, Oregon, neighborhoods. SARS-CoV-2 RNA was sequenced from all positive wastewater and nasal swab samples. Clinically reported case counts were obtained from the Oregon Health Authority.

RESULTS

Estimated community SARS-CoV-2 prevalence ranged from 8 to 1,687/10,000 persons. Community wastewater SARS-CoV-2 concentrations ranged from 2.9 to 5.1     log 10 gene copies per liter. Wastewater SARS-CoV-2 concentrations were more highly correlated (Pearson's r = 0.96 ; R 2 = 0.91 ) with community prevalence than were clinically reported cases of COVID-19 (Pearson's r = 0.85 ; R 2 = 0.73 ). Monte Carlo simulations indicated that wastewater SARS-CoV-2 concentrations were significantly better than clinically reported cases at estimating prevalence (p < 0.05 ). In addition, wastewater analyses determined neighborhood-level COVID-19 hot spots and identified SARS-CoV-2 variants (B.1 and B.1.399) at the neighborhood and city scales.

DISCUSSION

The greater reliability of wastewater SARS-CoV-2 concentrations over clinically reported case counts was likely due to systematic biases that affect reported case counts, including variations in access to testing and underreporting of asymptomatic cases. With these advantages, combined with scalability and low costs, wastewater-based epidemiology can be a key component in public health surveillance of COVID-19 and other communicable infections. https://doi.org/10.1289/EHP10289.

Cross-sectional study of COVID-19 knowledge, beliefs and prevention behaviours among adults in Senegal

OBJECTIVES

The aim of the study was to explore COVID-19 beliefs and prevention behaviours in a francophone West African nation, Senegal.

DESIGN

This was a cross-sectional analysis of survey data collected via a multimodal observational study.

PARTICIPANTS

Senegalese adults aged 18 years or older (n=1452).

PRIMARY AND SECONDARY OUTCOME MEASURES

Primary outcome measures were COVID-19 prevention behaviours. Secondary outcome measures included COVID-19 knowledge and beliefs. Univariate, bivariate and multivariate statistics were generated to describe the sample and explore potential correlations.

SETTING

Participants from Senegal were recruited online and telephonically between June and August 2020.

RESULTS

Mask wearing, hand washing and use of hand sanitiser were most frequently reported. Social distancing and staying at home were also reported although to a lower degree. Knowledge and perceived risk of COVID-19 were very high in general, but risk was a stronger and more influential predictor of COVID-19 prevention behaviours. Men, compared with women, had lower odds (adjusted OR (aOR)=0.59, 95% CI 0.46 to 0.75, p<0.001) of reporting prevention behaviours. Rural residents (vs urban; aOR=1.49, 95% CI 1.12 to 1.98, p=0.001) and participants with at least a high school education (vs less than high school education; aOR=1.33, 95% CI 1.01 to 1.76, p=0.006) were more likely to report COVID-19 prevention behaviours.

CONCLUSIONS

In Senegal, we observed high compliance with recommended COVID-19 prevention behaviours among our sample of respondents, in particular for masking and personal hygiene practice. We also identified a range of psychosocial and demographic predictors for COVID-19 prevention behaviours such as knowledge and perceived risk. Stakeholders and decision makers in Senegal and across Africa can use place-based evidence like ours to address COVID-19 risk factors and intervene effectively with policies and programming. Use of both phone and online surveys enhances representation and study generalisability and should be considered in future research with hard-to-reach populations.

Health literacy and hypertension outcomes in a multi-ethnic population: the HELIUS study

BACKGROUND

Hypertension disproportionately affects ethnic minority groups. Although health literacy may play role in these ethnic inequalities, little is known about the extent to which health literacy affects hypertension prevalence, awareness, treatment and control in different ethnic groups. Therefore, we assessed these associations in a multi-ethnic population.

METHODS

Baseline data from the HELIUS study were used including participants of Dutch (n = 1948), South-Asian Surinamese (n = 2054) and African Surinamese (n = 1932) origin aged 18-70 years, who lived in Amsterdam, the Netherlands, were fluent in Dutch and underwent health literacy assessment through the Rapid Estimate of Adult Literacy in Medicine-Dutch (REALM-D). The REALM-D was categorized either as low (<60 sumscore) or adequate (≥60 sumscore) health literacy. Participants completed questionnaires and underwent physical examination.

RESULTS

After adjusting for confounding variables, Dutch [odds ratio (OR) 2.02; 95% confidence interval (CI), 1.11-3.64] and African Surinamese (OR 1.36; 1.03-1.79) with low health literacy were more likely than those with adequate health literacy to have hypertension, whereas in South-Asian Surinamese this association was not significant. No significant associations were found between health literacy and hypertension awareness, treatment and control in any of the ethnic groups.

CONCLUSION

Findings indicate that health literacy is associated with hypertension prevalence in selected ethnic groups, but not with hypertension awareness, treatment and control. Targeting health literacy might be an entry point for tackling ethnic inequalities in hypertension prevalence. To substantially reduce these inequalities, further research is needed to explore other factors and pathways through which health literacy may impact hypertension outcomes in different ethnic groups.

Noise annoyance predicts symptoms of depression, anxiety and sleep disturbance 5 years later. Findings from the Gutenberg Health Study

BACKGROUND

Cross-sectional studies have shown that noise annoyance is strongly associated with mental distress, however, its long-term effects on mental health is unknown. We therefore investigated whether noise annoyance predicts depression, anxiety and sleep disturbance in a large, representative sample 5 years later.

METHODS

We investigated longitudinal data of N = 11 905 participants of the Gutenberg Health Study, a population-based, prospective, single-centre cohort study in mid-Germany (age at baseline 35-74 years). Noise annoyance was assessed at baseline and 5-year follow-up (sources: road traffic, aircraft, railways, industrial, neighbourhood indoor and outdoor noise; and day vs. nighttime). Depression, anxiety and sleep disturbance were assessed using the Patient Health Questionnaire-9 and Generalized Anxiety Disorder-2. Participants suffering from depression, anxiety or sleep disturbance at baseline were excluded from the respective multivariate analyses of new onset at follow-up.

RESULTS

General noise annoyance remained stable. Daytime noise annoyance predicted new onset of depressive, anxiety symptoms (also nighttime annoyance) and sleep disturbance (beyond respective baseline scores). Additional predictors were female sex, lower age and low socioeconomic status (SES). Regarding specific sources, daytime baseline aircraft annoyance predicted depression and anxiety. Sleep disturbance was most consistently predicted by neighbourhood annoyance (baseline and follow-up) and follow-up annoyance by aircraft (night) and road traffic (day and night).

CONCLUSIONS

We identified current and past noise annoyances as risk factors for mental distress and sleep disturbance. Furthermore, women, younger adults and those with lower SES are particularly susceptible to noise annoyance. Our results indicate the need to provide regulatory measures in affected areas to prevent mental health problems.

A feasibility study to assess Imbrasia belina (mopane worm) sensitisation and related respiratory health outcomes in a rural community in Gwanda district, Zimbabwe

Background

Allergic diseases are considered to be some of the fastest growing chronic conditions in Africa. Of concern is the paucity of knowledge about the local environment and its role in allergic disease development. In response to this, we explored whether Imbrasia belina, a popular indigenous edible insect commonly known as mopane worm, is a potential allergen of clinical and public health significance in Zimbabwe. This study was intended to assess the plausibility and feasibility of this hypothesis with a view to evaluate the insect’s health impact in a larger study.

Methods

The study participants included male and female villagers aged 10 years and above in Gwanda district, Zimbabwe. Eligible participants who completed the household questionnaire were referred to the local clinic for skin prick tests and to measure lung function and allergic airway inflammation. Allergen sensitisation patterns were evaluated using 10 different inhalant allergen extracts including an in-house preparation of mopane worm. Lung function was measured with a Koko Legend spirometer, and fractional exhaled nitric oxide levels (FeNO) (NIOX VERO) were measured in participants with at least one abnormal spirometric parameter. Data was analysed using Stata version 13 software.

Results

Of the 46 eligible participants that completed the household questionnaire, 17 went to the clinic giving a response rate of 37%. The majority who completed the questionnaire were adults (91%) and the children (9%) were all female. The prevalence of sensitisation to Imbrasia belina was 50%, and the prevalence ranged from 22 to 72% for the other allergens including cockroach, mosquito and house dust mites. The data collection tools were safe and well tolerated by participants with no adverse events reported. Self-reported respiratory symptoms, abnormal lung function and elevated FeNO were recorded amongst participants sensitised to mopane worm.

Conclusion

Pre-defined feasibility criteria were met with the exception of a lower than expected response rate for clinic data collection in this pilot study. For the main study, modifying the sampling strategy and applying more consistent community engagement will improve the response rates.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40814-021-00780-9.

Do Diabetes Mellitus Differences Exist within Generations? Three Generations of Moluccans in the Netherlands

Background: Diabetes mellitus (DM) is known to be more prevalent among migrants compared to their host populations. It is unclear whether DM prevalence differs between generations among migrants. We investigated the differences in DM prevalence among three generations of Moluccans, who have been living for over 65 years in the Netherlands, compared to the Dutch population. Methods: In this cross-sectional study, data of a healthcare insurance database on hospital and medication use (Achmea Health Database) were used. The dataset contained 5394 Moluccans and 52,880 Dutch persons of all ages. DM differences were assessed by means of logistic regression, adjusting for age, sex, urbanization, and area socio-economic status. Results: The prevalence of DM was higher in all generations of Moluccans compared to the Dutch. The adjusted odds ratios (AORs) for DM were significantly higher in total group of Moluccans compared to the Dutch (AOR 1.60, 95% CI 1.42–1.80) and across the first and second generation of Moluccans compared to the Dutch (first generation (1.73, 1.47–2.04) and second generation (1.44, 1.19–1.75). Higher AOR were found for first generation men (1.55, 1.22–1.97) and first (1.90, 1.52–2.37) and second (1.63, 1.24–2.13) generation Moluccan women compared to the Dutch. AOR for the third generation Moluccans was increased to a similar extent (1.51, 0.97–2.34), although not statistical significant. Conclusions: Our findings show higher odds of DM across generations of Moluccans compared to the Dutch. DM prevention strategies for minorities should be targeted at all migrant generations in host countries.

Healthcare Utilisation of Moluccans in the Netherlands: Equal Care for Equal Need after 60 Years of Residence in the Host Country?

Background: In many countries, recent migrants have difficulties using healthcare to the same extent as host populations. It is uncertain whether these differences persist for long-settled migrants. This study examined healthcare utilisation of Moluccans in 2012, more than 60 years after they migrated from Indonesia to the Netherlands. Methods: A survey was held among 715 Moluccans and 3417 Dutch persons. Differences in healthcare utilisation were assessed using regression analyses adjusting for age, gender, indicators of health, religious affiliation, and education. Results: Moluccans had lower rates of healthcare use, including visits to the general practitioner (odds ratio (OR) = 0.67), outpatient medical specialist (OR = 0.50), dentist (OR = 0.65), and physiotherapist (OR = 0.56), as well as the use of paid housekeeping services (OR = 0.37). Among those who visited a healthcare service, no difference was found between Moluccans and Dutch in the frequency of visits, except for physiotherapist visits (rate ratio (RR) = 0.51). For the risk of hospitalisation, no difference was found; however, of those admitted to the hospital, the frequency of admission was lower among Moluccans than Dutch (RR = 0.74). Conclusions: Despite their long residence in the host country, equal utilisation of healthcare services has not been achieved for Moluccans in the Netherlands. Demand-based factors (e.g., family networks, health beliefs, and use of traditional medicine) may contribute to the persistence of such differences and require further investigation.

Guided internet interventions for depression: impact of sociodemographic factors on treatment outcome in Indonesia

Depression is the leading cause of disability worldwide, but an alarming treatment gap exists, especially in lower- and middle income countries (LMIC), where people are exposed to many societal and sociodemographic risk factors. As internet access increases in LMIC, online interventions could decrease this gap, especially when shown suitable for all demographics, including vulnerable groups with low socioeconomic status (SES). We used mixed-model analysis to explore moderating effects of sociodemographic factors (age, sex, education level, SES and urbanicity) on treatment effect in a recent trial in Indonesia, comparing guided online behavioral activation versus online psychoeducation only for depression, in 313 participants from (sub)urban areas. Outcome measures were self-reported Patient Health Questionnaire 9 (PHQ-9) and Inventory of Depressive Symptomatology (IDS-SR). Without correction for multiple testing, we found urbanicity to moderate treatment effect, with stronger treatment effect in suburban relative to urban participants (IDS-SR 24 weeks past baseline, p = 0.04) and a trend towards moderation by SES, with stronger treatment effect in low SES groups (PHQ-9 10 weeks past baseline, p = 0.07). These exploratory results suggest online treatments are a promising mental health intervention for all demographics in a (sub)urban LMIC setting, but hypothesis-testing studies including rural participants are warranted.

A population-based retrospective study comparing cancer mortality between Moluccan migrants and the general Dutch population: equal risk 65 years after immigration?

OBJECTIVE

To test the hypothesis that cancer mortality rates among the Moluccan-Dutch, the oldest non-Western migrant group to arrive in the Netherlands after the Second World War, are similar to those in the general Dutch population.

DESIGN

Population-based retrospective study.

SETTING

Data from the national cause of death registry in the Netherlands and municipal registries.

PARTICIPANTS

Using historic records containing family names of all Moluccan-Dutch who arrived in the Netherlands in 1951, we identified 81 591 Moluccan-Dutch persons in the national cause of death registry of the Netherlands. The reference group consisted of 15 866 538 persons of the general Dutch population.

OUTCOME MEASURES

Mortality data were linked to demographic data from municipal registries. We calculated all-cancer and cancer-specific mortality and measured differences between the two groups using Poisson regression, adjusting for sex, age and area socioeconomic status. We conducted a sub-analysis for the first-generation and second-generation Moluccan-Dutch.

RESULTS

There was no difference in all-cancer mortality between Moluccan-Dutch and the general Dutch population. Mortality was higher among Moluccan-Dutch for liver, cervix and corpus uteri cancers, but lower for stomach, oesophagus, kidney and nervous system cancers. For most cancers, mortality risk as compared with the general Dutch population varied between different generations of Moluccan-Dutch.

CONCLUSIONS

Several decades after migration, the Moluccan-Dutch show similar all-cancer mortality, but different cancer-specific mortality rates, when compared with the general Dutch population.

All-cause mortality among three generations of Moluccans in the Netherlands

BACKGROUND

Mortality variations have been reported for ethnic minority groups compared with their host populations. It is uncertain how this mortality variations change over time as younger generations become older and mix with the host population. This study aimed to examine all-cause mortality among three generations of Moluccans in the Netherlands, and to compare Moluccans with a mixed and non-mixed ethnic background.

METHODS

We used data from the death and municipality registry for the years 2000 through 2013, including all registered Dutch inhabitants. A list of Moluccan surnames was used to select the Moluccan population. Mortality differences were calculated by Poisson regression, controlling for sex, age and district socio-economic status.

RESULTS

High all-cause mortality rates were observed in all generations of Moluccans although the extent of the differences between Moluccans and the Dutch were smaller in second (1.15, 1.07-1.23) and third generation (1.14, 1.00-1.29) compared with the first generation (1.55, 1.49-1.60). Higher all-cause mortality is also reflected in the higher mortality from most causes of death except neoplasms and external causes. Both mixed and non-mixed Moluccans showed high all-cause mortality among the first (child) and second generation compared with the Dutch.

CONCLUSION

Our findings show a higher all-cause mortality in three generations of Moluccans compared with the Dutch. The results show that mortality inequalities may persist, though in an attenuated form, over generations among ethnic minorities.

Recruitment of ethnic minorities for public health research: An interpretive synthesis of experiences from six interlinked Danish studies

Aims: This paper examines the importance of recruitment site in relation to the recruitment of ethnic minorities into health research. It presents a synthesis of experiences drawn from six interlinked Danish studies which applied different methods and used healthcare facilities and educational settings as sites for recruitment. Methods: Inspired by interpretive reviewing, data on recruitment methods from the different studies were synthesized with a focus on the various levels of recruitment success achieved. This involved an iterative process of comparison, analysis and discussion of experiences among the researchers involved. Results: Success in recruitment seemed to depend partly on recruitment site. Using healthcare facilities as the recruitment site and healthcare professionals as gatekeepers was less efficient than using schools and employees from educational institutions. Successful study designs also depended on the possibility of singling out specific locations with a high proportion of the relevant ethnic minority target population. Conclusions:The findings, though based on a small number of cases, indicate that health professionals and healthcare institutions, despite their interest in high-quality health research into all population groups, fail to facilitate research access to some of the most disadvantaged groups, who need to be included in order to understand the mechanisms behind health disparities. This happens despite the genuine wish of many healthcare professionals to help facilitate such research. In this way, the findings indirectly emphasize the specific challenge of accessing more vulnerable and sick groups in research studies.