Using Quantile Regression to Estimate Intervention Effects Beyond the Mean

Konstantopoulos, Spyros; Li, Wei; Miller, Shazia Rafiullah; Ploeg, Arie van der

doi:10.1177/0013164419837321

Cited by 37 publications

(23 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method of estimation was chosen because the assumptions of regression analysis were not met. Quantile regression can estimate the conditional median (i.e., 50 th quantile) on the outcome variable and makes no assumption regarding the distribution of the outcome [ 85 , 86 ]. In our analysis, each of the four groups of pandemic restrictions was treated as an outcome variable, whereas the four subdimensions of the PBQ (religious fundamentalism, xenophobia, acceptance of capitalism, and anti-welfare) served as predictor variables.…”

Section: Resultsmentioning

confidence: 99%

Political beliefs and the acceptance of the SARS-CoV-2 pandemic restrictions. The case of Poland

Turska-Kawa

Pilch

2022

PLoS ONE

View full text Add to dashboard Cite

We investigated the relationships between political beliefs regarding two aspects of the right-left distinction (cultural and economic) and the acceptance of the pandemic restrictions using variable-centred and person-centred approaches. The community sample consisted of 305 participants. Four groups of the restrictions were considered. Religious fundamentalism predicted positively the acceptance of the restrictions associated with the limitations of labour rights and those limiting civil rights without a direct impact on safety. Anti-welfare negatively predicted the acceptance of the restrictions regarding social distancing and those limiting civil rights and increasing safety. These associations were discussed in relation to basic needs and values which motivate persons who endorse right-wing or left-wing political views. The latent profile analysis revealed three profiles of political beliefs, which were termed “Conservative Statists,” “Liberal Laissez-fairists,” and “Conservative Laissez-fairists.” The profiles differed in terms of acceptance of the pandemic restrictions, and the patterns of these relationships were different for particular groups of restrictions.

show abstract

Section: Resultsmentioning

confidence: 99%

Political beliefs and the acceptance of the SARS-CoV-2 pandemic restrictions. The case of Poland

Turska-Kawa

Pilch

2022

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…With a multilevel stratified sampling method, we selected our participants from different geographic areas of China, covering both urban and rural areas. Second, in this study, we used QR modeling methodology instead of ordinary linear squares (OLS) regression, and thus, the results are more robust and comprehensive than only the mean estimates which are commonly more sensitive to outliers and can be influenced by imbalances of extreme data in the upper or lower tails of the distribution of outcome ( 43 ). Last but not the least, the quality of data collection guaranteed the reliability of health estimates in the first place.…”

Section: Discussionmentioning

confidence: 99%

The Effect of Body Adiposity and Alcohol Consumption on Serum Uric Acid: A Quantile Regression Analysis Based on the China National Health Survey

Pan²,

Ren³

et al. 2022

Front. Nutr.

View full text Add to dashboard Cite

Adiposity and alcohol consumption are reported to be associated with a higher level of serum uric acid (SUA), but whether their effect differs on SUA percentile distribution is still unclear. In this study, we aimed to investigate how alcohol intake and body fat percentage (%BF) integrated with body mass index (BMI) influence the distribution of SUA in Chinese adults. Data from the China National Health Survey (CNHS) which included adults from 10 provinces of China were used (n = 31,746, aged 20–80 years, 40% male). %BF and BMI were integrated into eight expanded body composition groups to understand how excess body adiposity affects the distribution of SUA in the populational level. Self-report alcohol intake information was collected by face-to-face questionnaire interview. Quantile regression (QR) was used to analyze the data. We found that adiposity and alcohol consumption were associated with SUA, especially at the upper percentile in both sexes. In obese men, the QR coefficients at the 75th and 95th percentiles were 74.0 (63.1–84.9) and 80.9 (52.5–109.3) μmol/L, respectively. The highest quartile of %BF in men had a 92.6 (79.3–105.9) μmol/L higher SUA levels at its 95th percentile than the 5th quartile (p < 0.001). Compared with normal or underweight with the lowest %BF group (NWBF1), the obesity-highest %BF group (OBBF4) had the strongest positive effect on SUA, especially at the higher percentile of SUA. In BMI-defined normal or underweight participants, a higher quartile of %BF had greater effect size in all SUA percentiles. In men, current alcohol drinking had the strongest effect at the 95th percentile of SUA (QR coefficient: 31.8, with 95% CI: 22.6–41.0) comparing with 14.5, 95% CI of 8.4 to 20.6 in the 5th SUA percentile. High risk of alcohol consumption had a greater effect on SUA, especially in the higher SUA percentile. The observation of stronger association at the higher percentile of SUA suggests that decreasing body adiposity and alcohol intake at the populational level may shift the upper tails of the SUA distributions to lower values, thereby reducing the incidence of hyperuricemia.

show abstract

“…For the wait time analysis, we compared the median wait times preimplementation and postimplementation in a difference-indifferences analysis. 14 Difference-in-differences analysis allows for measurement of intervention effects and background changes by comparing the pre-post trends in the intervention group with the pre-post trends in a comparison group. The method assumes that if the intervention did not make a difference, the change in the outcome of interest would be the same across the groups.…”

Section: Quantitative Methodsmentioning

confidence: 99%

eConsults' Impact on Care Access and Wait Times in Rheumatology

et al. 2022

View full text Add to dashboard Cite

Background/ObjectiveA growing number of health systems have implemented eConsults to improve access to specialty advice, but few studies have described their use in rheumatology or impact on visit wait times. We evaluated the uptake of an eConsult program and its impact on wait times for in-person rheumatology visits.MethodsIn this quality improvement project, we analyzed electronic health record data from 4 intervention clinics and 4 comparison clinics, 12 months before and after implementation of an eConsult program. We compared median wait time for rheumatology appointments using a pre-post difference-in-differences analysis and quantile regression, adjusting for patient age, race, sex, clinic pair, and primary insurance payer. We also interviewed 11 primary care providers from the intervention clinics and conducted a rheumatology provider focus group (n = 4) to elucidate experiences with the program.ResultsRheumatologists recommended management in primary care or referral to another specialty for 41% of eConsults, reducing initial demand for in-person visits. The median wait times dropped in the intervention and the comparison clinics (42 and 25 days, respectively). Intervention clinic median wait time dropped 17 days more than comparison clinics, and this was nonstatistically significant (p = 0.089). eConsults fit provider care tasks best for triage or initial workup for diagnosis, and less well when tests required interpretation, or when back and forth communication was needed to manage the patient's condition.ConclusionsImplementation of eConsults for rheumatology was associated with reduced wait times for rheumatology appointments and supported primary care providers in the triage and workup for a substantial portion of patients.

show abstract

Using Quantile Regression to Estimate Intervention Effects Beyond the Mean

Cited by 37 publications

References 21 publications

Political beliefs and the acceptance of the SARS-CoV-2 pandemic restrictions. The case of Poland

Political beliefs and the acceptance of the SARS-CoV-2 pandemic restrictions. The case of Poland

The Effect of Body Adiposity and Alcohol Consumption on Serum Uric Acid: A Quantile Regression Analysis Based on the China National Health Survey

eConsults' Impact on Care Access and Wait Times in Rheumatology

Contact Info

Product

Resources

About