Nonresponse data in sexual well-being among breast reconstruction patients-who are we overlooking?

BACKGROUND

Missing data can affect the representativeness and accuracy of survey results, and sexual health-related surveys are especially at a higher risk of nonresponse due to their sensitive nature and stigma. The purpose of this study was to evaluate the proportion of patients who do not complete the BREAST-Q Sexual Well-being relative to other BREAST-Q modules and compare responders versus nonresponders of Sexual Well-being. We secondarily examined variables associated with Sexual Well-being at 1-year.

METHODS

A retrospective analysis of patients who underwent breast reconstruction from January 2018 to December 2021 and completed any of the BREAST-Q modules postoperatively at 1-year was performed.

RESULTS

The 2941 patients were included. Of the four BREAST-Q domains, Sexual Well-being had the highest rate of nonresponse (47%). Patients who were separated (vs. married, OR = 0.69), whose primary language was not English (vs. English, OR = 0.60), and had Medicaid insurance (vs. commercial, OR = 0.67) were significantly less likely to complete the Sexual Well-being. Postmenopausal patients were significantly more likely to complete the survey than premenopausal patients. Lastly, autologous reconstruction patients were 2.93 times more likely to respond than implant-based reconstruction patients (p < 0.001) while delayed (vs. immediate, OR = 0.70, p = 0.022) and unilateral (vs. bilateral, OR = 0.80, p = 0.008) reconstruction patients were less likely to respond. History of psychiatric diagnosis, aromatase inhibitors, and immediate breast reconstruction were significantly associated with lower Sexual Well-being at 1-year.

CONCLUSION

Sexual Well-being is the least frequently completed BREAST-Q domain, and there are demographic and clinical differences between responders and nonresponders. We encourage providers to recognize patterns in nonresponse data for Sexual-Well-being to ensure that certain patient population's sexual health concerns are not overlooked.

Refractory Thyroid Eye Disease Unresponsive to Teprotumumab: A Case Report

Thyroid eye disease (TED) is a complex autoimmune condition that can cause proptosis, ophthalmoplegia, diplopia, optic nerve compression, and vision loss. These clinical findings are caused by a complex pathological mechanism characterized by thyroid-stimulating hormone receptor autoantibodies activating thyroid-stimulating hormone receptors (TSH-Rs). Overexpressed insulin-like growth factor 1 (IGF-1) receptors found in orbital fibroblasts form complexes with these TSH-Rs, leading to the inflammation and expansion of these tissues. Teprotumumab, a human monoclonal antibody sold under the brand name Tepezza, is currently the only FDA-approved immunotherapy for the treatment of TED. Given as an intravenous infusion every three weeks, teprotumumab works by suppressing IGF-1 receptors, thereby interfering with TSH-R and IGF-1 complex-mediated actions in these fibroblasts. The efficacy of teprotumumab was established in randomized, placebo-controlled clinical trials, which demonstrated clinically meaningful improvements in proptosis, inflammation, and diplopia. While teprotumumab has been shown to be efficacious, our patient with TSHRAb-positive euthyroid thyroid-associated ophthalmopathy who presented with diplopia did not have any significant improvement following the standard treatment dose of eight infusions over a 24-week period. This case underscores not only barriers to treatment, such as the high cost of teprotumumab but also highlights the importance of identifying risks for nonresponse.

Analysis of Interview Breakoff in the Behavioral Risk Factor Surveillance System, 2018 and 2019

Introduction

Survey breakoff is an important source of total survey error. Most studies of breakoff have been of web surveys-less is known about telephone surveys. In the past decade, the breakoff rate has increased in the Behavioral Risk Factor Surveillance System, the world's largest annual telephone survey. Analysis of breakoff in Behavioral Risk Factor Surveillance System can improve the quality of Behavioral Risk Factor Surveillance System. It will also provide evidence in research of total survey error on telephone surveys.

Methods

We used data recorded as breakoff in the 2018 and 2019 Behavioral Risk Factor Surveillance System. We converted questions and modules to a time variable and applied Kaplan-Meier method and a proportional hazard model to estimate the conditional and cumulative probabilities of breakoff and study the potential risk factors associated with breakoff.

Results

Cumulative probability of breakoffs up to the end of the core questionnaire was 7.03% in 2018 and 9.56% in 2019. The highest conditional probability of breakoffs in the core was 2.85% for the physical activity section. Cumulative probability of breakoffs up to the end of the core was higher among those states that inserted their own questions or optional modules than among those that did not in both years. The median risk ratio of breakoff among all states was 5.70 in 2018 and 3.01 in 2019. Survey breakoff was associated with the length of the questionnaire, the extent of expected recollection, and the location of questions.

Conclusions

Breakoff is not an ignorable component of total survey error and should be considered in Behavioral Risk Factor Surveillance System data analyses when variables have higher breakoff rates.

Which Outreach Modes Improve Response Rates to Physician Surveys? Lessons from an Experiment at the American Board of Internal Medicine

Physicians are a notoriously difficult group to survey due to a low propensity to respond. We investigate the relative effectiveness of reminder phone calls, pre-notification postcards, mailed paper surveys, and $1 upfront incentives for boosting survey response rate by embedding a randomized experiment into a mixed-mode operational survey at the American Board of Internal Medicine in 2019. Expected response rates and average marginal effects for each follow-up method were computed from a logistic regression model. The control group which only received email reminders achieved a response rate of 18.2%, 95% CI: (15.0%, 21.9%). The intervention group which included reminder emails, pre-notification postcards, and mailed paper surveys with $1 incentives achieved a response rate of 43.1%, 95% CI: (38.8%, 47.5%). Mailed paper surveys yielded the largest percentage point increase in response rate of 11.2%, 95% CI: (7.3%, 15.2%), while $1 upfront monetary incentives and phone call reminders increased survey response rate by 5.9%, 95% CI: (1.6%, 10.2%) and 5.5%, 95% CI: (2.6%, 8.3%) respectively. Pre-notification postcards are associated with a 2.0%, 95% CI: (-1.7%, 5.6%) increase in survey response rate. Cost-effectiveness for each method is discussed. This research supports optimal decision making for researchers when planning a physician survey study.

What They Are Not Telling Us: Analysis of Nonresponders on a National Survey of Resident Well-Being

To characterize nonrespondents to a national survey about trainee well-being, examine response patterns to questions of sensitive nature, and assess how nonresponse biases prevalence estimates of mistreatment and well-being.

Background

Surgical trainees are at risk for burnout and mistreatment, which are discernible only by self-report. Therefore, prevalence estimates may be biased by nonresponse.

Methods

A survey was administered with the 2018 and 2019 American Board of Surgery In-Training Examinations assessing demographics, dissatisfaction with education and career, mistreatment, burnout, thoughts of attrition, and suicidality. Responders in 2019 were characterized as survey "Completers," "Discontinuers" (quit before the end), and "Selective Responders" (selectively answered questions throughout). Multivariable logistic regression assessed associations of respondent type with mistreatment and well-being outcomes, adjusting for individual and program characteristics. Longitudinal survey identifiers linked survey responses for eligible trainees between 2018 and 2019 surveys to further inform nonresponse patterns.

Results

In 2019, 6956 (85.6%) of 8129 eligible trainees initiated the survey, with 66.5% Completers, 17.5% Discontinuers, and 16.0% Selective Responders. Items with the highest response rates included dissatisfaction with education and career (93.2%), burnout (86.3%), thoughts of attrition (90.8%), and suicidality (94.4%). Discontinuers and Selective Responders were more often junior residents and racially/ethnically minoritized than Completers. No differences were seen in burnout and suicidality rates between Discontinuers, Selective Responders, and Completers. Non-White or Hispanic residents were more likely to skip questions about racial/ethnic discrimination than non-Hispanic White residents (21.2% vs 15.8%; odds ratio [OR], 1.35; 95% confidence interval [CI], 1.19-1.53), particularly when asked to identify the source. Women were not more likely to omit questions regarding gender/gender identity/sexual orientation discrimination (OR, 0.91; 95% CI, 0.79-1.04) or its sources (OR, 1.02; 95% CI, 0.89-1.16). Both Discontinuers and Selective Responders more frequently reported physical abuse (2.5% vs 1.1%; P = 0.001) and racial discrimination (18.3% vs 13.6%; P < 0.001) on the previous survey (2018) than Completers.

Conclusions

Overall response rates are high for this survey. Prevalence estimates of burnout, suicidality, and gender discrimination are likely minimally impacted by nonresponse. Nonresponse to survey items about racial/ethnic discrimination by racially/ethnically minoritized residents likely results in underestimation of this type of mistreatment.

Predictors of nonresponse to dupilumab in patients with atopic dermatitis: A machine learning analysis

BACKGROUND

Many patients with atopic dermatitis (AD) have a suboptimal response to systemic therapy.

OBJECTIVE

This study assessed predictors of nonresponse to dupilumab in patients with AD.

METHODS

Data (April 2017 through June 2019) for patients aged ≥12 years with AD (ICD-9/10-CM: 691.8/L20.x) who initiated dupilumab on or after April 1, 2017 (index date) were collected from an electronic health record and insurance claims database. Nonresponse indicators (dupilumab discontinuation, addition of another systemic therapy or phototherapy, addition of a high-potency topical corticosteroid, AD-related hospital visit, AD-related emergency room visit, incident skin infection) were predicted from available demographic and clinical variables using machine learning.

RESULTS

Among 419 patients (mean age: 45 years), 145 (35%) experienced ≥1 indicator of nonresponse in the 6-month post-index period. In patients with ≥1 indicator, the most common was dupilumab discontinuation (47% [68/145]). Of note, this analysis could not capture nonmedical reasons of dupilumab discontinuation (eg, cost, access). The most common predictors of nonresponse were a claim for ibuprofen (in 69% of patients with a nonresponse indicator) and Quan-Charlson Comorbidity Index value of 3-4 (59%).

CONCLUSION

Systemic dupilumab therapy for AD can be associated with a relatively high prevalence of nonresponse indicators. Factors associated with these indicators -ie, predictors of nonresponse- may be used to optimize disease management.

Biomarkers to Predict DMARDs Efficacy and Adverse Effect in Rheumatoid Arthritis

Rheumatoid arthritis (RA), one of the most common immune system diseases, mainly affects middle-aged and elderly individuals and has a serious impact on the quality of life of patients. Pain and disability caused by RA are significant symptoms negatively affecting patients, and they are especially seen when inappropriate treatment is administered. Effective therapeutic strategies have evolved over the past few decades, with many new disease-modifying antirheumatic drugs (DMARDs) being used in the clinic. Owing to the breakthrough in the treatment of RA, the symptoms of patients who could not be treated effectively in the past few years have been relieved. However, some patients complain about symptoms that have not been reported, implying that there are still some limitations in the RA treatment and evaluation system. In recent years, biomarkers, an effective means of diagnosing and evaluating the condition of patients with RA, have gradually been used in clinical practice to evaluate the therapeutic effect of RA, which is constantly being improved for accurate application of treatment in patients with RA. In this article, we summarize a series of biomarkers that may be helpful in evaluating the therapeutic effect and improving the efficiency of clinical treatment for RA. These efforts may also encourage researchers to devote more time and resources to the study and application of biomarkers, resulting in a new evaluation system that will reduce the inappropriate use of DMARDs, as well as patients’ physical pain and financial burden.

Absence of early mood improvement as a robust predictor of rTMS nonresponse in major depressive disorder

BACKGROUND

Symptoms of major depressive disorder (MDD) are reported to change early in treatment with repetitive transcranial magnetic stimulation (rTMS). We evaluated early changes in sleep, anxiety, and mood as predictors of nonresponse to rTMS treatment.

METHODS

Three hundred twenty-nine subjects with nonpsychotic MDD completed a 6-week course of rTMS treatment. Subjects were stratified by the severity of their baseline depression, and had their overall depressive symptoms recorded every week of treatment. We evaluated lack of improvement in sleep, anxiety, and mood symptoms after 1 and 2 weeks as potential predictors of eventual nonresponse, defined as <50% improvement in compositive depressive symptoms after 6 weeks. This was measured as negative predictive value (NPV; the likelihood that lack of early symptom improvement accurately predicted eventual treatment nonresponse).

RESULTS

Subjects with severe or very severe baseline depression achieving <20% improvement in mood at 1 week were correctly predicted as nonresponders with NPVs largely >90%. At 2 weeks, subjects with very severe baseline depression who failed to demonstrate any improvement in mood were all nonresponders. Lack of improvement in sleep at 2 weeks was also a significant predictor.

CONCLUSIONS

Identifying a lack of early mood improvement is a practical and robust method to predict rTMS nonresponse. This suggests a treatment protocol change may be indicated in patients with more severe baseline depression showing minimal early mood improvement.

Unit Response and Costs in Web Versus Face-To-Face Data Collection: Comparison of Two Cross-sectional Health Surveys

Background

Potential is seen in web data collection for population health surveys due to its combined cost-effectiveness, implementation ease, and increased internet penetration. Nonetheless, web modes may lead to lower and more selective unit response than traditional modes, and this may increase bias in the measured indicators.

Objective

This research assesses the unit response and costs of a web study versus face-to-face (F2F) study.

Methods

Alongside the Belgian Health Interview Survey by F2F edition 2018 (BHISF2F; net sample used: 3316), a web survey (Belgian Health Interview Survey by Web [BHISWEB]; net sample used: 1010) was organized. Sociodemographic data on invited individuals was obtained from the national register and census linkages. Unit response rates considering the different sampling probabilities of both surveys were calculated. Logistic regression analyses examined the association between mode system and sociodemographic characteristics for unit nonresponse. The costs per completed web questionnaire were compared with the costs for a completed F2F questionnaire.

Results

The unit response rate is lower in BHISWEB (18.0%) versus BHISF2F (43.1%). A lower response rate was observed for the web survey among all sociodemographic groups, but the difference was higher among people aged 65 years and older (15.4% vs 45.1%), lower educated people (10.9% vs 38.0%), people with a non-Belgian European nationality (11.4% vs 40.7%), people with a non-European nationality (7.2% vs 38.0%), people living alone (12.6% vs 40.5%), and people living in the Brussels-Capital (12.2% vs 41.8%) region. The sociodemographic characteristics associated with nonresponse are not the same in the 2 studies. Having another European (OR 1.60, 95% CI 1.20-2.13) or non-European nationality (OR 2.57, 95% CI 1.79-3.70) compared to a Belgian nationality and living in the Brussels-Capital (OR 1.72, 95% CI 1.41-2.10) or Walloon (OR 1.47, 95% CI 1.15-1.87) regions compared to the Flemish region are associated with a higher nonresponse only in the BHISWEB study. In BHISF2F, younger people (OR 1.31, 95% CI 1.11-1.54) are more likely to be nonrespondents than older people, and this was not the case in BHISWEB. In both studies, lower educated people have a higher probability of being nonrespondent, but this effect is more pronounced in BHISWEB (low vs high education level: Web, OR 2.71, 95% CI 2.21-3.39 and F2F OR 1.70, 95% CI 1.48-1.95). The BHISWEB study had a considerable advantage; the cost per completed questionnaire was almost 3 times lower (€41 [US $48]) compared with F2F data collection (€111 [US $131]).

Conclusions

The F2F unit response rate was generally higher, yet for certain groups the difference between web and F2F was more limited. Web data collection has a considerable cost advantage. It is therefore worth experimenting with adaptive mixed-mode designs to optimize financial resources without increasing selection bias (eg, only inviting sociodemographic groups who are keener to participate online for web surveys while continuing to focus on increasing F2F response rates for other groups).

The mRNA Expression and Regulation of Glucocorticoid Receptor Isoforms Associated with Response of Multiple Myeloma Treated with a Glucocorticoids-Dependent Regimen of Dexamethasone, Bortezomib and Thalidomide

OBJECTIVE

Glucocorticoids (GCs) are the effective first-line drugs and indispensable in chemotherapy regimens to treat patients with multiple myeloma (MM). Previous studies in a variety of hematologic malignancies have shown that the biological action of GC is mediated through the expression and activation and of glucocorticoids receptor (GR) isoforms in vitro. GR and its regulation are crucial determinants of the efficacy of GC independent therapy. There is currently lack of research on patients with MM.

METHODS

132 patients with MM were divided into responders (78 cases) and nonresponders (54 cases) according to the efficacy evaluated after four cycles of GC-dependent regimen. 66 patients with iron-deficiency anemia were served as controls. Preparation of mononuclear bone marrow cells (MBMCs) was purified by Ficoll-Hypaque gradient centrifugation. The mRNA expression of GR α, β, γ, P, SRp30, SRp40, HSP90, NF-κB and AP-1 were detected by real time RT-PCR.

TRIAL REGISTRATION

CHiCTR-RCH-12002872.

RESULTS

The expression of four GR isoforms exhibited the following trend in MM patients and controls: GRα>GR-P>GRγ>GRβ. GRα and HSP90 expression in responders was significantly higher than that of the nonresponders (P<0.050). HSP90/GRα expression in MM patients exhibited significantly higher than that in controls (P<0.001). SRp30c and SRp40 mRNA expression both showed significant positive correlation with GRα transcript (P<0.001). Compared with controls, NF-kB and AP -1 expression in MM patients was higher. NF-kB and AP-1 expression of nonresponders were significantly higher than that of responders. The difference was not obvious statistically (P>0.050).

CONCLUSION

Our findings raise the possibility that low expression of GRα and HSP90 plays important roles in nonresponders. Lack of HSP90 might affect GR structure and further take part in nonresponse. SRp30c and SRp40 mRNA expression both showed significant positive correlation with GRα. That might become new targets for treatment of nonresponders in MM patients, although further studies are needed for clarification.

Participation Bias in a Survey of Community Patients With Heart Failure

OBJECTIVE

To identify differences between participants and nonparticipants in a survey of physical and psychosocial aspects of health among a population-based sample of patients with heart failure (HF).

PATIENTS AND METHODS

Residents from 11 Minnesota counties with a first-ever code for HF (International Classification of Diseases, Ninth Revision 428 and Tenth Revision I50) between January 1, 2013, and December 31, 2016, were identified. Participants completed a questionnaire by mail or telephone. Characteristics and outcomes were extracted from medical records and compared between participants and nonparticipants. Response rate was calculated using guidelines of the American Association for Public Opinion Research. The association between nonparticipation and outcomes was examined using Cox proportional hazards regression for death and Andersen-Gill modeling for hospitalizations.

RESULTS

Among 7911 patients, 3438 responded to the survey (American Association for Public Opinion Research response rate calculated using formula 2 = 43%). Clinical and demographic differences between participants and nonparticipants were noted, particularly for education, marital status, and neuropsychiatric conditions. After a mean ± SD of 1.5±1.0 years after survey administration, 1575 deaths and 5857 hospitalizations occurred. Nonparticipation was associated with a 2-fold increased risk for death (hazard ratio, 2.29; 95% CI, 2.05-2.56) and 11% increased risk for hospitalization (hazard ratio, 1.11; 95% CI, 1.02-1.22) after adjusting for age, sex, time from HF diagnosis to index date, marital status, coronary disease, arrhythmia, hyperlipidemia, diabetes, cancer, chronic kidney disease, arthritis, osteoporosis, depression, and anxiety.

CONCLUSION

In a large survey of patients with HF, participation was associated with notable differences in clinical and demographic characteristics and outcomes. Examining the impact of participation is critical to draw inference from studies of patient-reported measures.

Accuracy of Five Multiple Imputation Methods in Estimating Prevalence of Type 2 Diabetes based on STEPS Surveys

Background: This study was aimed to evaluate five Multiple Imputation (MI) methods in the context of STEP-wise Approach to Surveillance (STEPS) surveys.

Methods: We selected a complete subsample of STEPS survey data set and devised an experimental design consisted of 45 states (3 × 3 × 5), which differed by rate of simulated missing data, variable transformation, and MI method. In each state, the process of simulation of missing data and then MI were repeated 50 times. Evaluation was based on Relative Bias (RB) as well as five other measurements that were averaged over 50 repetitions.

Results: In estimation of mean, Predictive Mean Matching (PMM) and Multiple Imputation by Chained Equation (MICE) could compensate for the nonresponse bias. Ln and Box–Cox (BC) transformation should be applied when the nonresponse rate reaches 40% and 60%, respectively. In estimation of proportion, PMM, MICE, bootstrap expectation maximization algorithm (BEM), and linear regression accompanied by BC transformation could correct for the nonresponse bias. Our findings show that even with 60% of nonresponse rate some of the MI methods could satisfactorily result in estimates with negligible RB.

Conclusion: Decision on MI method and variable transformation should be taken with caution. It is not possible to regard one method as totally the worst or the best and each method could outperform the others if it is applied in its right situation. Even in a certain situation, one method could be the best in terms of validity but the other method could be the best in terms of precision.

The Effects of an Incentive Boost on Response Rates, Fieldwork Effort, and Costs across Two Waves of a Panel Study

This paper describes the association between an incentive boost and data collection outcomes across two waves of a long-running panel study. In a recent wave, with the aim of achieving response rate goals, all remaining sample members were offered a substantial incentive increase in the final weeks of data collection, despite uncertainty about potential effects on fieldwork outcomes in the following wave. The analyses examine response rates and the average number of interviewer attempts to complete the interview in the waves during and after the incentive boost, and provide an estimate of the cost of the incentives and fieldwork in the waves during and following the boost. The findings provide suggestive evidence that the use of variable incentive strategies from one wave to the next in the context of an ongoing panel study may be an effective strategy to reduce nonresponse and may yield enduring positive effects on subsequent data collection outcomes.

Using Administrative Data to Explore the Effect of Survey Nonresponse in the UK Employment Retention and Advancement Demonstration

BACKGROUND

Even a well-designed randomized control trial (RCT) study can produce ambiguous results. This article highlights a case in which full sample results from a large-scale RCT in the United Kingdom differ from results for a subsample of survey respondents.

OBJECTIVES

Our objective is to ascertain the source of the discrepancy in inferences across data sources and, in doing so, to highlight important threats to the reliability of the causal conclusions derived from even the strongest research designs.

RESEARCH DESIGN

The study analyzes administrative data to shed light on the source of the differences between the estimates. We explore the extent to which heterogeneous treatment impacts and survey nonresponse might explain these differences. We suggest checks which assess the external validity of survey measured impacts, which in turn provides an opportunity to test the effectiveness of different weighting schemes to remove bias. The subjects included 6,787 individuals who participated in a large-scale social policy experiment.

RESULTS

Our results were not definitive but suggest nonresponse bias is the main source of the inconsistent findings.

CONCLUSIONS

The results caution against overconfidence in drawing conclusions from RCTs and highlight the need for great care to be taken in data collection and analysis. Particularly, given the modest size of impacts expected in most RCTs, small discrepancies in data sources can alter the results. Survey data remain important as a source of information on outcomes not recorded in administrative data. However, linking survey and administrative data is strongly recommended whenever possible.

Ethnicity and response to primary three-dose hepatitis B vaccination in employees in the Netherlands, 1983 through 2017

BACKGROUND AND AIMS

Hepatitis B virus (HBV) vaccination is recommended to all employees who have an occupational risk in the Netherlands. This study assessed the determinants of the immune response to primary standard three-dose HBV vaccination (0, 1, 6 months), with the main focus on ethnicity.

METHODS

Out of 76 239 individuals who received HBV vaccination between April 1983 and December 2017, 11 567 persons with a known country of birth and complete vaccination schedule were included in this study. Weighted multiple logistic regression with Firth's bias adjustment was used to assess the determinants of nonresponse (anti-HBs < 10 mIU/mL) and low response (anti-HBs 10-99 mIU/mL).

RESULTS

Baseline characteristics of the study population (n = 11 567) were as follows: mean age 27.5 years (95% confidence interval [CI], 27.23-27.72), 99.4% born in the Netherlands and 93.5% of Western European origin. Of all identified subjects, 180 (1.6%) were HBV vaccine nonresponders and 549 (4.8%) were low responders. When compared with individuals aged <40 years, the rate of nonresponse (4.3% vs 0.8%; P < .001) and low response (11.9% vs 2.9%; P < .001) was higher in those aged 40 years or older. The height of anti-HBs levels were lower in those subjects aged >40 years in comparison with those younger than 40 years, P < .001. All nonresponders were born in the Netherlands. Although no significant association was found between nonresponse and individuals of Western European origin (adjusted odds ratio [aOR] = 1.20; 95% CI, 0.66-2.44; P = .163), low response to HBV vaccination was significantly associated with Western European origin (aOR = 2.21; 95% CI, 1.41-3.86; P = .001). Significant determinants for nonresponse were older age at vaccination (aOR = 1.06; 95% CI, 1.06-1.07; P < .001) and male gender (aOR = 2.51; 95% CI, 1.97-3.22; P < .001).

CONCLUSIONS

The nonresponse rate was low in our study population. Our findings suggest that the vaccines being used for the primary vaccination are probably less immunogenic for older individuals, males, and persons of Western European origin.

Tree-based Machine Learning Methods for Survey Research

Predictive modeling methods from the field of machine learning have become a popular tool across various disciplines for exploring and analyzing diverse data. These methods often do not require specific prior knowledge about the functional form of the relationship under study and are able to adapt to complex non-linear and non-additive interrelations between the outcome and its predictors while focusing specifically on prediction performance. This modeling perspective is beginning to be adopted by survey researchers in order to adjust or improve various aspects of data collection and/or survey management. To facilitate this strand of research, this paper (1) provides an introduction to prominent tree-based machine learning methods, (2) reviews and discusses previous and (potential) prospective applications of tree-based supervised learning in survey research, and (3) exemplifies the usage of these techniques in the context of modeling and predicting nonresponse in panel surveys.

How Many Patients With Schizophrenia Do Not Respond to Antipsychotic Drugs in the Short Term? An Analysis Based on Individual Patient Data From Randomized Controlled Trials

OBJECTIVE

An important clinical question is how many patients with acute schizophrenia do not respond to antipsychotics despite being treated for adequate time and with an effective dose. However, up to date, the exact extent of the phenomenon remains unclear.

METHODS

We calculated the nonresponse and nonremission percentages using individual patient data from 16 randomized controlled trials (RCTs). Six thousand two hundred twenty-one patients were assigned to one antipsychotic (amisulpride, flupenthixol, haloperidol, olanzapine, quetiapine, risperidone, or ziprasidone) at an adequate dose; the response was assessed at 4-6 weeks. As various definitions of nonresponse have been used in the literature, we applied 4 different cut-offs covering the whole range of percent Positive and Negative Syndrome Scale (PANSS)/Brief Psychiatric Rating Scale (BPRS) reduction (≤0%, <25%, <50%, <75%).For symptomatic remission, we used the definition proposed by Andreasen without employing the time criterion.

RESULTS

The overall nonresponse for the cut-off of ≤0% PANSS/BPRS reduction was 19.8% (18.8%-20.8%); for the cut-off of <25% reduction it was 43% (41.7%-44.3%); for the cut-off of <50% reduction it was 66.5% (65.3%-67.8%); and for the cut-off of <75% reduction it was 87% (86%-87.9%). The overall percentage of no symptomatic remission was 66.9% (65.7%-68.1%). Earlier onset of illness, lower baseline severity and the antipsychotic used were significantly associated with higher nonresponse percentages.

CONCLUSIONS

Nonresponse and nonremission percentages were notably high. Nevertheless, the patients in our analysis could represent a negative selection since they came from short-term RCTs and could have been treated before study inclusion; thus, further response may not have been observed. Observational studies on this important question are needed.

Using Administrative Data to Explore the Effect of Survey Nonresponse in the UK Employment Retention and Advancement Demonstration

BACKGROUND

Even a well-designed randomized control trial (RCT) study can produce ambiguous results. This article highlights a case in which full sample results from a large-scale RCT in the United Kingdom differ from results for a subsample of survey respondents.

OBJECTIVES

Our objective is to ascertain the source of the discrepancy in inferences across data sources and, in doing so, to highlight important threats to the reliability of the causal conclusions derived from even the strongest research designs.

RESEARCH DESIGN

The study analyzes administrative data to shed light on the source of the differences between the estimates. We explore the extent to which heterogeneous treatment impacts and survey nonresponse might explain these differences. We suggest checks which assess the external validity of survey measured impacts, which in turn provides an opportunity to test the effectiveness of different weighting schemes to remove bias. The subjects included 6,787 individuals who participated in a large-scale social policy experiment.

RESULTS

Our results were not definitive but suggest nonresponse bias is the main source of the inconsistent findings.

CONCLUSIONS

The results caution against overconfidence in drawing conclusions from RCTs and highlight the need for great care to be taken in data collection and analysis. Particularly, given the modest size of impacts expected in most RCTs, small discrepancies in data sources can alter the results. Survey data remain important as a source of information on outcomes not recorded in administrative data. However, linking survey and administrative data is strongly recommended whenever possible.

Effects of Caregiver and Decedent Characteristics on CAHPS Hospice Survey Scores

CONTEXT

The Consumer Assessment of Healthcare Providers and Systems (CAHPS^®) Hospice Survey assesses the care experiences of hospice patients and their families. Public reporting of hospice performance on these survey measures began in February 2018.

OBJECTIVES

Develop an appropriate case-mix adjustment (CMA) model to allow for fair comparisons between hospices.

METHODS

We analyzed CAHPS Hospice Survey data reflecting experiences of 915,442 patients who received care from 2513 hospice programs between April 2015 and March 2016. Decedent and caregiver characteristics were identified for inclusion in CMA based on their variation across hospices (as measured by intraclass correlation coefficients [ICCs]) and how predictive they were of responses to survey questions (as assessed by linear regression).

RESULTS

The final CMA model included decedent age, payer for hospice care, primary diagnosis, length of final episode of hospice care, caregiver age, caregiver education, relationship to decedent, survey language/language spoken at home, and response percentile. The characteristics that varied most across hospices were language (ICC = 0.48 for Spanish survey or home language) and payer for hospice care (ICC = 0.42 for Medicare only; ICC = 0.35 for Medicare and private insurance). The characteristics that were most predictive of caregivers' survey responses were payer for hospice care, caregiver education, and survey language/language spoken at home. Lack of appropriate adjustment would incorrectly rank hospices by 1.2-5.4 percentile points.

CONCLUSION

To ensure fair comparisons across hospices, CAHPS Hospice Survey measure scores should be adjusted for several caregiver and decedent characteristics.

Variability in Individual Response to Aerobic Exercise Interventions Among Older Adults

Although a plethora of evidence supports the benefits of exercise among older adults, a majority of studies have emphasized group differences, while giving little, if any, attention to individual differences. Given the lack of data on variability in response, the present review examined how nonresponse to aerobic exercise has been defined in older adult populations and characteristics associated with nonresponse among older adults. The results of this review suggest that interindividual variability in response of maximal oxygen consumption to aerobic exercise interventions is prevalent among older adults (1.4-63.4%); age, sex, race, and body mass index may not be critical determinants of nonresponse; whereas health status, baseline fitness, and exercise dose appear important. Future intervention studies should evaluate and report the variability in individual response of older adults to exercise; investigators should develop programs that allow for modification of components to assist older adults in achieving optimal benefit from exercise programs.

Non-response to sad mood induction: implications for emotion research

Experimental induction of sad mood states is a mainstay of laboratory research on affect and cognition, mood regulation, and mood disorders. Typically, the success of such mood manipulations is reported as a statistically significant pre- to post-induction change in the self-rated intensity of the target affect. The present commentary was motivated by an unexpected finding in one of our studies concerning the response rate to a well-validated sad mood induction. Using the customary statistical approach, we found a significant mean increase in self-rated sadness intensity with a moderate effect size, verifying the "success" of the mood induction. However, that "success" masked that, between one-fifth and about one-third of our samples (adolescents who had histories of childhood-onset major depressive disorder and healthy controls) reported absolutely no sadness in response to the mood induction procedure. We consider implications of our experience for emotion research by (1) commenting upon the typically overlooked phenomenon of nonresponse, (2) suggesting changes in reporting practices regarding mood induction success, and (3) outlining future directions to help scientists determine why some subjects do not respond to experimental mood induction.

Platelet function-guided modification in antiplatelet therapy after acute ischemic stroke is associated with clinical outcomes in patients with aspirin nonresponse

Purpose

To investigate the association of clinical outcomes with platelet function-guided modification in antiplatelet therapy in patients with ischemic stroke.

Results

Among 812 patients, 223 patients had aspirin nonresponse, 204 patients was modified in antiplatelet therapy after platelet function testing. Mean follow-up period was 4.8 ± 1.7 years (ranged from 1 to 6.4 years). The incidence rates of ischemic events, death, or bleeding events were not significantly different between the patients with and without antiplatelet therapy modification. However, in patients with aspirin nonresponse, antiplatelet therapy modification was associated with decreased ischemic events (hazard ratio, 0.67; 95% confidence interval [CI], 0.62–0.97; P = 0.01) and ischemic stroke (hazard ratio, 0.70; 95% CI, 0.63–0.98; P = 0.03) compared with no modification in antiplatelet therapy.

Conclusions

In patients with aspirin nonresponse, platelet function-guided modification in antiplatelet therapy after an ischemic stroke was associated with significantly lower rate of ischemic events. The platelet function testing may be useful to guide antiplatelet therapy modification.

Methods

This is a retrospective, multicentre study. From August 2010 to December 2014, 812 patients with ischemic stroke underwent platelet function testing using platelet aggregation. Antiplatelet therapy modification was defined as any change in antiplatelet therapy after testing, including increasing aspirin dosage, adding an additional antiplatelet agent to aspirin, or switching to a more potent antiplatelet agent. The primary outcome was ischemic events. Secondary outcomes included death and bleeding events. Clinical outcomes were compared between patients with and without antiplatelet therapy modification using univariate and propensity score-adjusted analyses.

Non-ignorable missingness in logistic regression

Nonresponses and missing data are common in observational studies. Ignoring or inadequately handling missing data may lead to biased parameter estimation, incorrect standard errors and, as a consequence, incorrect statistical inference and conclusions. We present a strategy for modelling non-ignorable missingness where the probability of nonresponse depends on the outcome. Using a simple case of logistic regression, we quantify the bias in regression estimates and show the observed likelihood is non-identifiable under non-ignorable missing data mechanism. We then adopt a selection model factorisation of the joint distribution as the basis for a sensitivity analysis to study changes in estimated parameters and the robustness of study conclusions against different assumptions. A Bayesian framework for model estimation is used as it provides a flexible approach for incorporating different missing data assumptions and conducting sensitivity analysis. Using simulated data, we explore the performance of the Bayesian selection model in correcting for bias in a logistic regression. We then implement our strategy using survey data from the 45 and Up Study to investigate factors associated with worsening health from the baseline to follow-up survey. Our findings have practical implications for the use of the 45 and Up Study data to answer important research questions relating to health and quality-of-life. Copyright © 2017 John Wiley & Sons, Ltd.

Relationship between Future Time Orientation and Item Nonresponse on Subjective Probability Questions: A Cross-Cultural Analysis

Time orientation is an unconscious yet fundamental cognitive process that provides a framework for organizing personal experiences in temporal categories of past, present and future, reflecting the relative emphasis given to these categories. Culture lies central to individuals' time orientation, leading to cultural variations in time orientation. For example, people from future-oriented cultures tend to emphasize the future and store information relevant for the future more than those from present- or past-oriented cultures. For survey questions that ask respondents to report expected probabilities of future events, this may translate into culture-specific question difficulties, manifested through systematically varying "I don't know" item nonresponse rates. This study drew on the time orientation theory and examined culture-specific nonresponse patterns on subjective probability questions using methodologically comparable population-based surveys from multiple countries. The results supported our hypothesis. Item nonresponse rates on these questions varied significantly in the way that future-orientation at the group as well as individual level was associated with lower nonresponse rates. This pattern did not apply to non-probability questions. Our study also suggested potential nonresponse bias. Examining culture-specific constructs, such as time orientation, as a framework for measurement mechanisms may contribute to improving cross-cultural research.

Malignant stroke in a ticagrelor non-responder as a complication following aneurysm treatment with the Pipeline Embolization Device™

The Pipeline Embolization Device™ (PED; Covidien Neurovascular Inc, Irvine, CA, USA) is a flow-diverting stent often used for the endovascular treatment of large or giant, wide-necked intracranial aneurysms of the internal carotid artery. Because of the inherent thrombogenicity of intracranial stents, dual-antiplatelet therapy is initiated after placement, which has been shown to decrease morbidity and mortality related to perioperative ischemic events in neurointerventional procedures. However, in some series, as much as 50% of patients demonstrate clopidogrel non-responsiveness. In these non-responders, alternate agents such as ticagrelor can be used to achieve adequate anticoagulation. Compared with clopidogrel, a prodrug requiring Cytochrome P450 enzymolysis for activation, ticagrelor directly and reversibly inhibits the P2Y12 ADP receptor. The absorption of the prodrug and the formation of its active metabolite is comparatively quicker ( t_max 1.3-2 hours; 1.5-3 hours, respectively). To date, there have been no documented cases of ticagrelor non-responsiveness involving patients undergoing placement of flow-diverting stents or other endovascular neuro-interventional procedures.

Atypical antipsychotics: recent research findings and applications to clinical practice: Proceedings of a symposium presented at the 29th Annual European College of Neuropsychopharmacology Congress, 19 September 2016, Vienna, Austria

Available evidence suggests that second-generation atypical antipsychotics are broadly similar to first-generation agents in terms of their efficacy, but may have a more favourable tolerability profile, primarily by being less likely to cause extrapyramidal symptoms. However, atypical antipsychotics are variably associated with disturbances in the cardiometabolic arena, including increased body weight and the development of metabolic syndrome, which may reflect differences in their receptor binding profiles. Effective management of schizophrenia must ensure that the physical health of patients is addressed together with their mental health. This should therefore involve consideration of the specific tolerability profiles of available agents and individualization of treatment to minimize the likelihood of adverse metabolic sequelae, thereby improving long-term adherence and optimizing overall treatment outcomes. Alongside this, modifiable risk factors (such as exercise, diet, obesity/body weight and smoking status) must be addressed, in order to optimize patients' overall health and quality of life (QoL). In addition to antipsychotic-induced side effects, the clinical management of early nonresponders and psychopharmacological approaches for patients with treatment-resistant schizophrenia remain important unmet needs. Evidence suggests that antipsychotic response starts early in the course of treatment and that early nonresponse accurately predicts nonresponse over the longer term. Early nonresponse therefore represents an important modifiable risk factor for poor efficacy and effectiveness outcomes, since switching or augmenting antipsychotic treatment in patients showing early nonresponse has been shown to improve the likelihood of subsequent treatment outcomes. Recent evidence has also demonstrated that patients showing early nonresponse to treatment with lurasidone at 2 weeks may benefit from an increase in dose at this timepoint without compromising tolerability/safety. However, further research is required to determine whether these findings are generalizable to other antipsychotic agents.

Estimating the Prevalence of Ovarian Cancer Symptoms in Women Aged 50 Years or Older: Problems and Possibilities

Diagnostic testing is recommended in women with "ovarian cancer symptoms." However, these symptoms are nonspecific. The ongoing Diagnosing Ovarian Cancer Early (DOVE) Study in Montreal, Quebec, Canada, provides diagnostic testing to women aged 50 years or older with symptoms lasting for more than 2 weeks and less than 1 year. The prevalence of ovarian cancer in DOVE is 10 times that of large screening trials, prompting us to estimate the prevalence of these symptoms in this population. We sent a questionnaire to 3,000 randomly sampled women in 2014-2015. Overall, 833 women responded; 81.5% reported at least 1 symptom, and 59.7% reported at least 1 symptom within the duration window specified in DOVE. We explored whether such high prevalence resulted from low survey response by applying inverse probability weighting to correct the estimates. Older women and those from deprived areas were less likely to respond, but only age was associated with symptom reporting. Prevalence was similar in early and late responders. Inverse probability weighting had a minimal impact on estimates, suggesting little evidence of nonresponse bias. This is the first study investigating symptoms that have proven to identify a subset of women with a high prevalence of ovarian cancer. However, the high frequency of symptoms warrants further refinements before symptom-triggered diagnostic testing can be implemented.

Do Differential Response Rates to Patient Surveys Between Organizations Lead to Unfair Performance Comparisons?: Evidence From the English Cancer Patient Experience Survey

BACKGROUND

Patient surveys typically have variable response rates between organizations, leading to concerns that such differences may affect the validity of performance comparisons.

OBJECTIVE

To explore the size and likely sources of associations between hospital-level survey response rates and patient experience.

RESEARCH DESIGN, SUBJECTS, AND MEASURES

Cross-sectional mail survey including 60 patient experience items sent to 101,771 cancer survivors recently treated by 158 English NHS hospitals. Age, sex, race/ethnicity, socioeconomic status, clinical diagnosis, hospital type, and region were available for respondents and nonrespondents.

RESULTS

The overall response rate was 67% (range, 39% to 77% between hospitals). Hospitals with higher response rates had higher scores for all items (Spearman correlation range, 0.03-0.44), particularly questions regarding hospital-level administrative processes, for example, procedure cancellations or medical note availability.From multivariable analysis, associations between individual patient experience and hospital-level response rates were statistically significant (P<0.05) for 53/59 analyzed questions, decreasing to 37/59 after adjusting for case-mix, and 25/59 after further adjusting for hospital-level characteristics.Predicting responses of nonrespondents, and re-estimating hypothetical hospital scores assuming a 100% response rate, we found that currently low performing hospitals would have attained even lower scores. Overall nationwide attainment would have decreased slightly to that currently observed.

CONCLUSIONS

Higher response rate hospitals have more positive experience scores, and this is only partly explained by patient case-mix. High response rates may be a marker of efficient hospital administration, and higher quality that should not, therefore, be adjusted away in public reporting. Although nonresponse may result in slightly overestimating overall national levels of performance, it does not appear to meaningfully bias comparisons of case-mix-adjusted hospital results.

Everolimus treatment for patients with autoimmune hepatitis and poor response to standard therapy and drug alternatives in use

OBJECTIVE

Not all patients with autoimmune hepatitis (AIH) respond to standard medical therapy with corticosteroids and azathioprine. Such patients may develop end-stage liver disease with poor prognosis unless transplantation is considered. Alternatively, the introduction of new therapeutic strategies could potentially ameliorate deterioration of liver function. Patients in our tertiary center were selected for everolimus therapy when exhibiting nonresponse or intolerance to combinations of the standard and empirical drugs in use (e.g., mycophenolate mofetil, calcineurin inhibitors [CNIs]). We here report the efficacy of everolimus treatment of patients with AIH.

MATERIALS AND METHODS

Seven patients (six female, mean age 47 years, range 22-62 years) in whom disease control could not be achieved with standard therapy or the alternative drugs in use were included.

RESULTS

Treatment with everolimus induced a clear reduction of transaminases within 2 weeks. After 3-5 months three patients had normal alanine aminotransferase (ALT) levels (10-45 IU) and four patients had ALT levels below 55 IU compared to a three- to fivefold elevated level prior to everolimus treatment. Sustained remission after 1 year of treatment was observed in three patients; in another two patients ALT was 45-68 U/L. Four patients in remission after 3 years were rebiopsied. Two showed no histological progression, and in two the fibrosis had decreased. Side effects noted were myalgias and minor bacterial infections not leading to discontinuation of the drug.

CONCLUSION

Our experience indicates that everolimus may be of value in selected patients with therapy-resistant AIH and comorbidity/side effects that excludes the use of CNIs.

Pretreatment and process predictors of nonresponse at different stages of inpatient psychotherapy

OBJECTIVE

Up to 50% of psychotherapeutic treatments end without significant improvements. While there is first evidence about predictors of nonresponse in outpatient psychotherapy, there are currently no studies investigating predictors of nonresponse in inpatient settings. Based upon a previous systematic literature review, we analyzed the predictive value of initial patient characteristics on nonresponse in symptom distress.

METHODS

Treatment episodes from 546 patients, treated for at least 4 weeks, were assessed under naturalistic conditions. Nonresponse status (i.e., lack of a reliable improvement in symptom distress) was investigated at four different time points: at week 4, at discharge, and at a two follow-ups (3 and 12 months after discharge). Hierarchical binary logistic regression models were used to predict nonresponse. Sociodemographic data, clinical variables, and the previous response status were entered subsequently in the model.

RESULTS

A moderate or functional level of initial symptom distress, a comorbid personality disorder, and previous nonresponse were the most consistent predictors of nonresponse.

CONCLUSIONS

The results point to the importance of early outcome assessment and suggest the implementation of more symptom-specific treatments.

The optimization of treatment and management of schizophrenia in Europe (OPTiMiSE) trial: rationale for its methodology and a review of the effectiveness of switching antipsychotics

BACKGROUND

Most of the 13 542 trials contained in the Cochrane Schizophrenia Group's register just tested the general efficacy of pharmacological or psychosocial interventions. Studies on the subsequent treatment steps, which are essential to guide clinicians, are largely missing. This knowledge gap leaves important questions unanswered. For example, when a first antipsychotic failed, is switching to another drug effective? And when should we use clozapine? The aim of this article is to review the efficacy of switching antipsychotics in case of nonresponse. We also present the European Commission sponsored "Optimization of Treatment and Management of Schizophrenia in Europe" (OPTiMiSE) trial which aims to provide a treatment algorithm for patients with a first episode of schizophrenia.

METHODS

We searched Pubmed (October 29, 2014) for randomized controlled trials (RCTs) that examined switching the drug in nonresponders to another antipsychotic. We described important methodological choices of the OPTiMiSE trial.

RESULTS

We found 10 RCTs on switching antipsychotic drugs. No trial was conclusive and none was concerned with first-episode schizophrenia. In OPTiMiSE, 500 first episode patients are treated with amisulpride for 4 weeks, followed by a 6-week double-blind RCT comparing continuation of amisulpride with switching to olanzapine and ultimately a 12-week clozapine treatment in nonremitters. A subsequent 1-year RCT validates psychosocial interventions to enhance adherence.

DISCUSSION

Current literature fails to provide basic guidance for the pharmacological treatment of schizophrenia. The OPTiMiSE trial is expected to provide a basis for clinical guidelines to treat patients with a first episode of schizophrenia.

Survey nonresponse among ethnic minorities in a national health survey--a mixed-method study of participation, barriers, and potentials

OBJECTIVES

The participation rate in the Danish National Health Survey (DNHS) 2010 was significantly lower among ethnic minorities than ethnic Danes. The purpose was to characterize nonresponse among ethnic minorities in DNHS, analyze variations in item nonresponse, and investigate barriers and incentives to participation.

DESIGN

This was a mixed-method study. Logistic regression was used to analyze nonresponse using data from DNHS (N = 177,639 and chi-square tests in item nonresponse analyses. We explored barriers and incentives regarding participation through focus groups and cognitive interviews. Informants included immigrants and their descendants of both sexes, with and without higher education.

RESULTS

The highest nonresponse rate was for non-Western descendants (80.0%) and immigrants 25 (72.3%) with basic education. Immigrants and descendants had higher odds ratios (OR = 3.07 and OR = 3.35, respectively) for nonresponse than ethnic Danes when adjusted for sex, age, marital status, and education. Non-Western immigrants had higher item nonresponse in several question categories. Barriers to non-participation related to the content, language, format, and layout of both the questionnaire and the cover letter. The sender and setting in which to receive the questionnaire also influenced answering incentives. We observed differences in barriers and incentives between immigrants and descendants.

CONCLUSIONS

Nonresponse appears related to linguistic and/or educational limitations, to alienation generated by the questions' focus on disease and cultural assumptions, or mistrust regarding anonymity. Ethnic minorities seem particularly affected by such barriers. To increase survey participation, questions could be sensitized to reflect multicultural traditions, and the impact of sender and setting considered.

Assessing the representativeness of population-sampled health surveys through linkage to administrative data on alcohol-related outcomes

Health surveys are an important resource for monitoring population health, but selective nonresponse may impede valid inference. This study aimed to assess nonresponse bias in a population-sampled health survey in Scotland, with a focus on alcohol-related outcomes. Nonresponse bias was assessed by examining whether rates of alcohol-related harm (i.e., hospitalization or death) and all-cause mortality among respondents to the Scottish Health Surveys (from 1995 to 2010) were equivalent to those in the general population, and whether the extent of any bias varied according to sociodemographic attributes or over time. Data from consenting respondents (aged 20–64 years) to 6 Scottish Health Surveys were confidentially linked to death and hospitalization records and compared with general population counterparts. Directly age-standardized incidence rates of alcohol-related harm and all-cause mortality were lower among Scottish Health Survey respondents compared with the general population. For all years combined, the survey-to-population rate ratios were 0.69 (95% confidence interval: 0.61, 0.76) for the incidence of alcohol-related harm and 0.89 (95% confidence interval: 0.83, 0.96) for all-cause mortality. Bias was more pronounced among persons residing in more deprived areas; limited evidence was found for regional or temporal variation. This suggests that corresponding underestimation of population rates of alcohol consumption is likely to be socially patterned.

Efficacy of lead-in silibinin and subsequent triple therapy in difficult-to-treat HIV/hepatitis C virus-coinfected patients

OBJECTIVES

The efficacy of current hepatitis C virus (HCV) triple therapy, including a protease inhibitor, is limited in HIV/HCV-coinfected patients with advanced liver fibrosis and nonresponse to previous peginterferon-ribavirin. These patients have a low chance (only 30%) of achieving a sustained virological response (SVR) during triple therapy and cannot wait for next-generation anti-HCV drugs. In a pilot study, we investigated the efficacy of a lead-in therapy with silibinin before triple therapy in difficult-to-treat patients.

METHODS

Inclusion criteria were HIV/HCV coinfection with advanced liver fibrosis and documented failure of previous peginterferon-ribavirin treatment. Intervention was lead-in therapy with intravenous silibinin 20 mg/kg/day for 14 days. Subsequently, peginterferon-ribavirin combined with telaprevir was initiated for 12 weeks, followed by peginterferon-ribavirin dual therapy until week 48 after initiation of triple therapy. The outcome measurements were HCV RNA after silibinin lead-in, at weeks 2, 4 and 12 of triple therapy, and SVR at week 24 after the end of treatment.

RESULTS

We examined six HIV/HCV-coinfected patients (four infected with genotype 1a). All had fibrosis grade METAVIR ≥F3 and were on fully suppressive antiretroviral therapy. Mean HCV RNA decline after silibinin therapy was 2.6 log10 IU/mL (range 2-3 log10 IU/mL). Five of the six patients were virologically suppressed at weeks 2 and 4, and all six at week 12 of triple therapy. One experienced a viral breakthrough thereafter. Four of five patients (80%) showed an SVR 24. One patient had an SVR 12 but has not yet reached week 24.

CONCLUSIONS

A lead-in with silibinin before triple therapy is highly effective and increases the probability of HCV treatment success in difficult-to-treat HIV/HCV-coinfected patients with advanced liver fibrosis and previous failure of peginterferon-ribavirin.

Extreme nonresponse to acute phase cognitive therapy for depression: an attempt to replicate and extend

As with other interventions for major depressive disorder (MDD), cognitive therapy (CT) results in treatment failure for about half of all participants. In 2007, Coffman and colleagues in Seattle studied this topic by identifying a group of patients who demonstrated an extremely poor response to CT (i.e., posttreatment BDI score≥31). They called these patients "extreme nonresponders" (ENR) and described the pretreatment characteristics that predicted response status. In the current study, we attempt a replication of the Seattle study with a larger sample of adults with recurrent MDD (N=473) who received a 16-20 session (12-14week) course of CT. The rate of ENR in this large sample was only 6.3% (30/473), compared to 22.2% (10/45) in the Seattle sample. Four pretreatment measures of symptom severity and functioning differed significantly among ENR and non-ENR participants. In each case, higher symptoms or poorer functioning were associated with ENR status. However, the combination of these factors in a regression model did not predict actual ENR status with the high degree of sensitivity or specificity observed in the Seattle study. These findings suggest that extreme nonresponse to CT is not as common as previously described and, although poor outcomes are associated with pretreatment clinical status, it is difficult to predict posttreatment symptom severity with a high degree of accuracy across different research samples.

Outcome for headache and pain-free nonresponders to treatment of the first attack: a pooled post-hoc analysis of four randomized trials of eletriptan 40 mg

OBJECTIVE

The objective of this article is to evaluate, in first attack eletriptan headache and pain-free nonresponders, the efficacy of treating a second and third attack with the same dose of eletriptan 40 mg (ELE-40).

METHODS

Data were pooled from four randomized, double-blind, placebo-controlled, multiple attack studies of eletriptan in the treatment of migraine. The first-attack eletriptan headache (HNR) and pain-free (PFNR) nonresponder samples consisted of patients who did not achieve headache or pain-free responses at two hours, or sustained headache or pain-free responses at 24 hours. The efficacy of the same dose of eletriptan (vs placebo; PBO) in treating the second and third attacks was evaluated using a logistic regression model.

RESULTS

Among Attack 1 eletriptan HNRs, treatment with ELE-40 (vs PBO) was associated with significantly higher two-hour headache response and pain-free rates, respectively, on both Attack 2 (48.8% vs 20.2%; 17.0% vs 3.9%; P  < 0.0001 for both comparisons) and Attack 3 (37.4% vs 15.5%; 18.8% vs 3.2%; P  < 0.0001 for both comparisons). Significantly higher sustained headache response and pain-free rates at 24 hours were also observed on both Attack 2 and Attack 3.

CONCLUSIONS

The results of this pooled analysis suggest that patients who have HNR or PFNR to an initial dose of eletriptan may respond when a second and third attack is treated with the same dose.

Methods for Handling Missing Secondary Respondent Data

Secondary respondent data are underutilized because researchers avoid using these data in the presence of substantial missing data. We reviewed, critically evaluated, and tested potential solutions to this problem. Five strategies of dealing with missing partner data are reviewed: complete case analysis, inverse probability weighting, correction with a Heckman selection model, maximum likelihood estimation, and multiple imputation. Two approaches were used to evaluate the performance of these methods. First, we used data from the National Survey of Fertility Barriers (N = 1,666) to estimate a model predicting marital quality based on characteristics of women and their husbands. Second, we conducted a simulation based on these data testing the five methods and compared the results to estimates where the true value was known. We found that the maximum likelihood and multiple imputation methods were advantageous because they allow researchers to utilize all of the available information as well as produce less biased and more efficient estimates.

The Effects of a Between-Wave Incentive Experiment on Contact Update and Production Outcomes in a Panel Study

Since 1969, families participating in the U.S. Panel Study of Income Dynamics (PSID) have been sent a mailing asking them to update or verify their contact information in order to keep track of their whereabouts between waves. Having updated contact information prior to data collection is associated with fewer call attempts, less tracking, and lower attrition. Based on these advantages, two experiments were designed to increase response rates to the between-wave contact mailing. The first experiment implemented a new protocol that increased the overall response rate by 7 - 10 percentage points compared to the protocol in place for decades on the PSID. This article provides results from the second experiment which examines the basic utility of the between-wave mailing, investigates how incentives affect article cooperation to the update request and field effort, and attempts to identify an optimal incentive amount. Recommendations for the use of contact update strategies in panel studies are made.

Nonparticipation in a Danish cohort study of long-term sickness absence

AIM

The aim of the present study was to identify predictors of nonparticipation in a Danish cohort of individuals on long-term sickness absence with a nonparticipation rate of 53.6%.

METHODS

Data from Danish public registers were linked to all 2414 individuals initially recruited to the cohort. Information regarding social- and health-related characteristics was retrieved. Adjusted logistic regression was carried out to examine differences between participants and nonparticipants as well as to identify predictors of nonparticipation.

RESULTS

Nonparticipation was associated with being male, relatively young, having a vocational secondary education, and having a low income, whereas a recent somatic disease treated in hospital was a predictor for participation. Having had a psychiatric disorder in the past was generally a barrier for participation, while a recent psychiatric disorder was a positive factor for participation.

CONCLUSION

Individuals with low socioeconomic status and individuals with prior psychiatric disorders were less willing to participate in this cohort study of long-term sickness absence.

Differing response to antipsychotic therapy in schizophrenia: pharmacogenomic aspects

Treatment-resistance in schizophrenia remains a public health problem: about 20% to 30% of patients do not respond to antipsychotic therapy. Clozapine has been shown to be effective in about one-third of patients, but the medical risks and weekly blood tests limit its broad application. While the heterogeneity of the disease and the duration of untreated psychosis are important, pharmacogenomic aspects must also be considered. Pharmacogenomic investigations offer the opportunity to individualize antipsychotic therapy according to the growing knowledge of the function and effect of the genetic polymorphisms that affect the pharmacokinetics and pharmacodynamics of antipsychotics. On the pharmacokinetic level, polymorphic phase I and II drug-metabolizing enzymes and transport proteins affect drug concentration at the target structure. The cytochrome P450 enzymes, N-acetyltransferase, and multidrug resistance protein (MDR1) particularly influence this parameter. Genetic alterations affecting drug pharmacodynamic properties have an impact on therapeutic outcome that is generally independent of the applied dosage regimen. A combined analysis of genetic polymorphisms in the dopaminergic and serotonergic receptors, neurotransmitter transporters, and other target structures involved in psychiatric disorders is already a powerful predictor of therapeutic outcome. An understanding of other factors influencing gene expression and protein production will facilitate individualized therapy in the future.

Association of host pharmacodynamic effects with virologic response to pegylated interferon alfa-2a/ribavirin in chronic hepatitis C

Patients receiving therapy for chronic hepatitis C virus (HCV) infection frequently experience cytopenias and weight loss. We retrospectively assessed the pharmacodynamic effects of pegylated interferon (PEG-IFN) alfa-2a and ribavirin by evaluating the relationship between changes in hematologic parameters, body weight, and virologic response. Patients with HCV genotypes 1, 4, 5, or 6 receiving 24 or 48 weeks of PEG-IFN alfa-2a and ribavirin therapy were pooled from four phase 3/4 trials. Maximum decreases in hemoglobin level, neutrophil count, platelet count, and weight during therapy were assessed according to virologic response category (sustained virologic response [SVR], relapse, breakthrough, and nonresponder) and race/ethnicity. Of 1,778 patients analyzed, more than half were male, non-Hispanic Caucasian, and infected with HCV genotype 1; had a baseline HCV RNA >800,000; and had alanine aminotransferase levels ≤3 × the upper limit of normal. Virologic responders (SVR, relapse, and breakthrough) experienced greater maximum decreases from baseline in hemoglobin level, neutrophil count, platelet count, and weight compared with nonresponders; however, no clear trend was observed between SVR, relapse, and breakthrough. After adjusting for drug exposure and treatment duration, only decreases in neutrophil count remained associated with virologic response. Significantly greater declines in neutrophil (P < 0.0001) and platelet (P < 0.005) count were observed at weeks 4, 12, and 24 of therapy in virologic responders compared with nonresponders. This difference between responders and nonresponders was also observed among racial/ethnic groups, although statistical significance was not consistent across all groups.

CONCLUSION

This post hoc analysis of HCV patients treated with PEG-IFN alfa-2a and ribavirin shows that maximum decreases from baseline in hematologic parameters and weight loss were associated with virologic response. However, after adjusting for drug exposure and accounting for duration of therapy, only neutropenia was independently associated with virologic response.

Bias Mechanisms in Intention-to-Treat Analysis With Data Subject to Treatment Noncompliance and Missing Outcomes

An analytical approach was employed to compare sensitivity of causal effect estimates with different assumptions on treatment noncompliance and non-response behaviors. The core of this approach is to fully clarify bias mechanisms of considered models and to connect these models based on common parameters. Focusing on intention-to-treat analysis, systematic model comparisons are performed on the basis of explicit bias mechanisms and connectivity between models. The method is applied to the Johns Hopkins school intervention trial, where assessment of the intention-to-treat effect on school children's mental health is likely to be affected by assumptions about intervention noncompliance and nonresponse at follow-up assessments. The example calls attention to the importance of focusing on each case in investigating relative sensitivity of causal effect estimates with different identifying assumptions, instead of pursuing a general conclusion that applies to every occasion.

Alternative methods for handling attrition: an illustration using data from the Fast Track evaluation

Using data from the evaluation of the Fast Track intervention, this article illustrates three methods for handling attrition. Multiple imputation and ignorable maximum likelihood estimation produce estimates that are similar to those based on listwise-deleted data. A panel selection model that allows for selective dropout reveals that highly aggressive boys accumulate in the treatment group over time and produces a larger estimate of treatment effect. In contrast, this model produces a smaller treatment effect for girls. The article's conclusion discusses the strengths and weaknesses of the alternative approaches and outlines ways in which researchers might improve their handling of attrition.