Training health care providers to administer VIA as a screening test for cervical cancer: a systematic review of essential training components.

BACKGROUND: Training health care providers to administer visual inspection after application of acetic acid (VIA) is paramount in improving cervical cancer screening services for women in low- and middle-income countries. The objective of this systematic review was to create a framework of essential VIA training components and provide illustrating examples of how VIA training programs can be carried out in different clinical settings. METHODS: A systematic review of PubMed, Embase, and Web of Science (from 2006 to 2021) was undertaken. Our inclusion criteria comprised articles reporting on implemented cervical cancer screening programs using VIA in a screen-and-treat approach. Trained health care providers with any level of health education were included, and the outcome of interest was the reporting of training components. Data were extracted by two reviewers, and a narrative synthesis of the training programs was performed. We developed a framework of seven essential training components and applied it to assess how training courses were conducted in different settings. RESULTS: 13 primary studies were eligible for inclusion, including 2,722 trained health care providers and 342,889 screened women. Most training courses lasted 5-7 days and included theoretical education, practical skill development, and competence assessment. It was unclear how visual aids and training in client counselling and quality assessment were integrated in the training courses. After the training course, nearly all the VIA training programs made provisions for on-job training at the providers' own clinical settings through supervision, feedback, and refresher training. CONCLUSIONS: This study demonstrates the feasibility of implementing international training recommendations for cervical cancer screening in real-world settings and provides valuable examples of training program implementation across various clinical settings. The diverse reporting practices of quality indicators in different studies hinder the establishment of direct links between these data and training program effectiveness. To enhance future reporting, authors should emphasize specific training components, delivery methods, and contextual factors. Standardized reporting of quality indicators for effective evaluation of VIA training programs is recommended, fostering comparability, facilitating research, and enhancing reporting quality in this field.

Are symptoms assessed differently for schizophrenia and other psychoses in legal insanity evaluations of violent crimes?

BACKGROUND: Forensic evaluations of legal insanity include the experts' assessment of symptoms present at the mental state examination (MSE) and the mental state at the time of offense (MSO). Delusions and hallucinations are most important. We explored how often symptoms were recorded in written forensic reports. DESIGN: This exploratory, cross-sectional study included 500 reports of legal insanity written in 2009-2018 from cases of violent crimes in Norway. The first author read all reports and coded symptoms recorded from the experts' assessments of the offenders. Two co-authors repeated this procedure for 50 randomly selected reports. Interrater reliability was calculated with Gwet's AC1. Generalized Linear Mixed Models with Wald tests for fixed effects and Risk Ratios as effect sizes were used for the statistical analyses. RESULTS: Legal insanity was the main conclusion in 23.6% of the reports; 71.2% of these were diagnosed with schizophrenia while 22.9% had other psychotic disorders. Experts recorded few symptoms from MSO, but more from MSE, although MSO is important for insanity. We found a significant association between delusions and hallucinations recorded present in the MSO and legal insanity for defendants with other psychotic disorders, but no association for defendants with schizophrenia. The differences in symptom recordings between diagnoses were significant. CONCLUSION: Few symptoms were recorded from the MSO. We found no association between presence of delusions or hallucinations and legal insanity for defendants with schizophrenia. This may indicate that a schizophrenia diagnosis is more important to the forensic conclusion than the symptoms recorded in the MSO.

Factor structure and internal consistency of the parent patient activation measure (P-PAM) in parents of children with ADHD in norwegian paediatric mental health.

BACKGROUND: This study aimed to explore the internal consistency and factor validity of the 13-item self-report questionnaire Parent-Patient Activation Measure (P-PAM) in a sample of parents of children with Attention-deficit/hyperactivity disorder. METHODS: In a cross-sectional study, 239 parents were recruited from four outpatient clinics of the Child and Adolescent Mental Health Services and completed the P-PAM along with demographic variables. The factor structure of the P-PAM was examined through exploratory factor analysis, and internal consistency was estimated with the use of both Cronbach's alpha and McDonald's omega. A confirmatory factor analysis was used to estimate and test individual parameters. RESULTS: The fit indices suggest an acceptable two-factor model of P-PAM and show high internal consistency and reliability for both factors, indicating that the scale measures two concepts. CONCLUSIONS: Our findings provide evidence for an acceptable factor structure and a high reliability of P-PAM as a measure of parent activation, suggesting that the theoretical factors reflect the construct of parent activation as intuitively compiled into an inner cognitive factor and an outer behavioral factor, which are related.

The safe pilot study: A prospective naturalistic study with repeated measures design to test protective factors against violence in and after discharge from forensic facilities.

Assessing violence risk amongst forensic patients is a vital legal and clinical task. The field of violence risk assessment has developed considerably over the past two decades but remains primarily risk focused. Despite this, growing attention to and work on protective factors or strengths has occurred. In this prospective naturalistic study with repeated observer-rated measures of 27 forensic patients, we tested the role of three potentially important but understudied dynamic protective factors: hope, insight, and resilience, along with a history of criminality, in terms of their impact on violence. Main effects models indicated that higher hopelessness and past criminal convictions were predictive of violence acts; higher resilience was associated with lower violence. In interaction models, hopelessness remained predictive. Importantly, there were significant interactions between resilience and past criminal convictions, with higher levels of resilience leading to lower violence, most amongst those with criminal convictions, and between resilience and hopelessness related emotional distress, in that higher resilience at high levels of patient acknowledged emotional distress due to hopelessness led to lower violence. Findings indicate the importance of focusing on strengths or protective factors in the assessment of risk and treatment planning for forensic patients. Despite the small sample, the repeated measures design was feasible and informative.

Use of assessment instruments in forensic evaluations of criminal responsibility in Norway.

OBJECTIVES: Assessment instruments are often used to enhance quality and objectivity in therapeutic and legal settings. We aimed to explore the use of instruments in Norwegian reports of forensic evaluations of criminal responsibility; specifically, whether this use was associated with diagnostic and forensic conclusions. METHODS: Our study has an exploratory cross-sectional design. We examined 500 reports filed with the Norwegian Board of Forensic Medicine in 2009-2018 regarding defendants indicted for the most serious violent crimes. The first author coded data from all reports according to a registration form developed for this study. Two co-authors then coded a random sample of 50 reports, and inter-rater reliability measures were calculated. The first author coded 41 reports for calculation of intra-rater reliability. Descriptive statistics are presented for the use of assessment instruments, and a generalized linear mixed model (GLMM) was used to estimate associations between the use of instruments and diagnostic and forensic conclusions. RESULTS: Instruments were used in 50.0% of reports. The Wechler's Adult Intelligence Scale (WAIS), Historical Clinical Risk-20 (HCR-20), and the Structured Clinical Interview for DSM disorders (SCID I), were used in 15.8, 13.8, and 9.0% of reports, respectively. The use of instruments increased from 36% in 2009 to 58% in 2015; then decreased to 49% in 2018. Teams of two experts wrote 98.0% of reports, and 43.4% of these teams comprised two psychiatrists. In 20.0% of reports, the diagnostic conclusion was schizophrenia, and in 8.8% it was other psychotic disorders. A conclusion of criminal irresponsibility was given in 25.8% of reports. Instruments were more often used in reports written by teams that comprised both a psychiatrist and a psychologist, compared to reports by two psychiatrists. The use of instruments was strongly associated with both diagnostic and forensic conclusions. CONCLUSION: Instruments were used in 50% of reports on forensic evaluations of criminal responsibility in Norway, and their use increased during the study period. Use of instruments was associated with diagnostic and forensic conclusions.

Lifestyle and Work-Related Factors Associated with Work Ability and Work Participation for People with Obesity: A Prospective Observational Study After Vocational Rehabilitation.

PURPOSE: We aimed to investigate which changes in the explanatory factors that were associated with positive change in the work ability score (WAS) and degree of work participation (DWP) for participants in a new 1-year vocational rehabilitation (VR) program for people on or at risk of sick leave due to obesity or obesity-related problems. PATIENTS AND METHODS: This prospective observational study included 95 participants with a body mass index (BMI) above 30 kg/m2. The 1-year multidisciplinary VR program with an integrated work and lifestyle intervention included 4 weeks of inpatient stay followed-up by five meetings. Differences between baseline and 12-month follow-up data were analyzed for the change in explanatory variables WAS, DWP, health-related quality of life (HRQoL), BMI, and return-to-work self-efficacy (RTWSE). The primary outcome was measured by multiple linear regression for predicting WAS and DWP. RESULTS: We found significant changes in WAS (1.51, 95% CI: 0.83 to 2.20, p<0.001), DWP (18.69, 95% CI: 8.35 to 29.02, p<0.001), HRQoL (2.57, 95% CI: 1.35 to 3.79, p<0.001), BMI (-2.33, 95% CI: -3.10 to -1.56, p<0.001), and in RTWSE (15.89, 95% CI: 4.07 to 27.71, p = 0.009). Regression analysis yielded a strong association between WAS at 12-month follow-up with an increase in HRQoL (ß=0.27, 95% CI: 0.16 to 0.38, p<0.001) and WAS baseline (ß=0.49, 95% CI: 0.28 to 0.71, p<0.001). Further, regression analysis demonstrated a strong association between DWP at 12-month follow-up with return-to-work expectancy (RTWEXP) (ß=-10.62, 95% CI: -15.25 to -6.03, p<0.001). CONCLUSION: The results indicate positive changes in WAS, DWP, HRQoL, BMI, and RTWSE from baseline to 12-month follow-up. For people with BMI above 30 kg/m2, changes in HRQoL are important for an increase in WAS, and a high RTWEXP is essential to achieve work participation. Future studies examining VR programs with lifestyle interventions for people with obesity are recommended.

Does Women's Screening History Have Any Impact on Mammography Screening Attendance After Tailored Education?: A Systematic Review and Meta-analysis.

BACKGROUND: Many ethnic minority women have low attendance at breast cancer screening. OBJECTIVES: This brief report explores whether women's screening histories impact mammography screening attendance after tailored education. RESEARCH DESIGN: Systematic searches were conducted in 5 databases. Randomized controlled trials of educational interventions tailored to ethnic minority women that measured attendance at mammography screening were eligible for inclusion. Data extraction and risk of bias assessment were performed independently. Data were combined in a meta-analysis by using random effects models. Heterogeneity was estimated by using I2 statistics. RESULTS: Six studies with 3521 women were eligible for inclusion. The D+L pooled risk ratio (RR) for mammography attendance for never screened participants was 1.54 (95% confidence interval, 1.24-1.91; P<0.001), with low heterogeneity (I2=27.1%, P=0.231). The D+L pooled risk ratio for attendance for ever screened participants was 1.26 (95% confidence interval, 1.11-1.43; P<0.001), with low heterogeneity (I2=35.5%, P=0.213). CONCLUSIONS: Tailored education increased attendance at mammography by 54% among never screened women and 26% among ever screened women. Although these findings must be interpreted with caution, the findings suggest that women's screening history is an important and ignored variable that affects how effective tailored education is on mammography screening attendance.

Changes to body mass index, work self-efficacy, health-related quality of life, and work participation in people with obesity after vocational rehabilitation: a prospective observational study.

BACKGROUND: People on or at risk of sick leave from work due to obesity or obesity-related problems participated in a new vocational rehabilitation (VR). The study aimed to examine the outcome changes in the participants' health-related quality of life (HRQoL), body mass index (BMI), return to work self-efficacy (RTWSE), work ability scale (WAS) and degree of work participation (DWP) after their participation in the 12-month VR programme. The secondary aim was to examine associations between the outcome changes and HRQoL at 12-month follow-up, measured with the HRQoL 15D instrument (15D). METHODS: This prospective observational study included 95 participants. The one-year multidisciplinary VR programme with an integrated work and lifestyle intervention included 4 weeks of inpatient stay followed-up by 5 meetings. A paired sample t-test was used to examine changes in HRQoL, BMI, RTWSE, WAS, and DWP between baseline and the 12-month follow-up. Multiple linear regression analyses explored associations between changes in HRQoL and the outcome variables. RESULTS: The participants achieved statistically significant changes in HRQoL (2.57, 95% CI: 1.35 to 3.79), BMI (- 2.33, 95% CI: - 3.10 to - 1.56), RTWSE (15.89, 95% CI: 4.07 to 27.71), WAS (1.51, 95% CI: 0.83 to 2.20) and DWP (18.69, 95% CI: 8.35 to 29.02). At 12 months, a significant association was found between HRQoL and BMI (B = - 0.34, 95% CI: - 0.65 to - 0.04), RTWSE (B = 0.02, 95% CI: 0.004 to 0.04), WAS (B = 0.91, 95% CI: 0.55 to 1.28), DWP (B = - 0.02, 95% CI: - 0.04 to 0.001) and work absence (B = - 0.01, 95% CI: - 0.02 to - 0.002). The regression model explained 71.8% of the HRQoL variance. CONCLUSION: The results indicated positive changes in HRQoL, BMI, RTWSE, WAS and DWP from baseline to the 12-month follow-up. Factors associated with HRQoL at the 12-month follow-up were decreased BMI, increased RTWSE, improved WAS and reduced work absence. Future studies examining VR programmes with lifestyle interventions for people with obesity are recommended. TRIAL REGISTRATION: Norwegian Regional Committee for Medical and Health Research Ethics (REC) 2017/573, Clinical Trials NCT03286374 , registered 18. September 2017. https://clinicaltrials.gov/ct2/results?cond=Obesity&term=Anita+Dyb+Linge&cntry=NO&state=&city=&dist=.

The Safe pilot study: A prospective naturalistic study with repeated measures design to test the psychosis - violence link in and after discharge from forensic facilities.

The research evidence is very strong for high recidivism rates of violence after discharge from forensic facilities. Big data research has found that a substantial proportion of the forensic population with relapse into violence has a psychosis diagnosis and a criminal record. However, more research on the association between psychotic symptoms and violence may inform and enhance risk assessment, prevention, and treatment. We conducted a prospective naturalistic study with a repeated measures design in a sample of 22 psychotic patients during follow-up after discharge from forensic mental health facilities. We had three aims: to test the predictive validity of three psychotic symptom scales for violence, to analyze main and interaction effects between psychotic symptoms and previous criminal conviction, and to explore the feasibility and potential benefit of the repeated measures design for prospective follow-up research. Interpreted within the limitation of the small sample size, the results were promising for all scales, particularly for adjusted effects without interaction. Two scales remained significant when their interaction with criminal conviction was adjusted. This indicates that risk judgment of psychotic patients with criminal conviction can be improved by adding measurement of fluctuations in psychotic symptoms. The repeated measures design was instrumental in this research.

Assessment of the Quality of Life in Parents of Children With ADHD: Validation of the Multicultural Quality of Life Index in Norwegian Pediatric Mental Health Settings.

Background: The brief generic Multicultural Quality of Life Index (MQLI) is a culturally informed self-report 10-item questionnaire used to measure health-related quality of life (QoL). QoL is an important outcome measure in guiding healthcare and is held as a substantial parameter to evaluate the effectiveness of healthcare. Attention Deficit Hyperactivity Disorder (ADHD) in children might negatively influence the parents' QoL. Having a validated questionnaire to measure QoL for this population will therefore be a vital first step in guiding healthcare for parents of children with ADHD. We aimed to examine the reliability and validity of the Norwegian version of the MQLI in a sample of parents of children with ADHD. Methods: In a cross-sectional study, 128 parents of children with ADHD were recruited from four outpatient clinics within the Child and Adolescents Mental Health Services (CAMHS) in Norway. They completed the MQLI along with an alternative well-being scale, the Five-item World Health Organization Well-being Index (WHO-5), and a form including demographic variables. Reliability and validity of the MQLI were examined. We conducted a factor analysis and calculated internal consistency and the correlation between the MQLI and the WHO-5. Results: Factor analysis of the parents reported MQLI yielded a one-factor solution. For the MQLI, Cronbach's alpha was 0.73. The correlation between the two measures of MQLI and WHO-5 was high (r = 0.84), reflecting convergent validity since the association between the two measures was strong. Conclusion: Results from this study support the reliability and validity of the Norwegian version of the MQLI for assessment of quality of life in parents of children with ADHD with good psychometric properties. Study findings support the use of the questionnaire in CAMHS.

Effect of culturally tailored education on attendance at mammography and the Papanicolaou test.

OBJECTIVES: To determine the effectiveness of culturally tailored education on attendance at breast and cervical cancer screening among ethnic minority women. DATA SOURCES: Systematic database searches in Ovid MEDLINE, ProQuest, PubMed, PsycINFO, and Cochrane CENTRAL. STUDY DESIGN: Randomized controlled trials (RCTs) of culturally tailored educational interventions to ethnic minority women in Western countries were investigated for a meta-analysis. RCTs that assessed attendance at mammography or the Papanicolaou test (Pap test) were eligible for inclusion. DATA COLLECTION METHODS: Study characteristics and results were extracted separately. Independent raters assessed risk of bias by using Cochrane Collaboration's tool. PRINCIPAL FINDINGS: Seven RCTs (n = 4246) were included in the meta-analysis of mammography attendance, and four RCTs (n = 1750) were included in the meta-analysis of Pap test attendance. The effect of culturally tailored educational interventions on attendance at mammography was an increase of 18 percent (RR = 1.18, 95% CI, 1.09-1.28, P < .001), with low heterogeneity (I2  = 30.0, P = .237), and a 54 percent increase at the Pap test (RR = 1.54, 95% CI, 1.14-2.09, P = .005), with substantial heterogeneity (I2  = 75.9%, P = .001). CONCLUSIONS: Interpreted within the limitations set by the low number of studies and substantial heterogeneity for the Pap test, findings from the current meta-analyses indicate that culturally tailored educational interventions may increase attendance of ethnic minority women at breast and cervical cancer screenings. There is a need for more studies, in particular RCTs conducted outside the United States, to determine if such findings are similar in other countries.

Observed to expected or logistic regression to identify hospitals with high or low 30-day mortality?

INTRODUCTION: A common quality indicator for monitoring and comparing hospitals is based on death within 30 days of admission. An important use is to determine whether a hospital has higher or lower mortality than other hospitals. Thus, the ability to identify such outliers correctly is essential. Two approaches for detection are: 1) calculating the ratio of observed to expected number of deaths (OE) per hospital and 2) including all hospitals in a logistic regression (LR) comparing each hospital to a form of average over all hospitals. The aim of this study was to compare OE and LR with respect to correctly identifying 30-day mortality outliers. Modifications of the methods, i.e., variance corrected approach of OE (OE-Faris), bias corrected LR (LR-Firth), and trimmed mean variants of LR and LR-Firth were also studied. MATERIALS AND METHODS: To study the properties of OE and LR and their variants, we performed a simulation study by generating patient data from hospitals with known outlier status (low mortality, high mortality, non-outlier). Data from simulated scenarios with varying number of hospitals, hospital volume, and mortality outlier status, were analysed by the different methods and compared by level of significance (ability to falsely claim an outlier) and power (ability to reveal an outlier). Moreover, administrative data for patients with acute myocardial infarction (AMI), stroke, and hip fracture from Norwegian hospitals for 2012-2014 were analysed. RESULTS: None of the methods achieved the nominal (test) level of significance for both low and high mortality outliers. For low mortality outliers, the levels of significance were increased four- to fivefold for OE and OE-Faris. For high mortality outliers, OE and OE-Faris, LR 25% trimmed and LR-Firth 10% and 25% trimmed maintained approximately the nominal level. The methods agreed with respect to outlier status for 94.1% of the AMI hospitals, 98.0% of the stroke, and 97.8% of the hip fracture hospitals. CONCLUSION: We recommend, on the balance, LR-Firth 10% or 25% trimmed for detection of both low and high mortality outliers.

Recent gestational diabetes was associated with mothers stopping predominant breastfeeding earlier in a multi-ethnic population.

AIM: It has previously been shown that breastfeeding may reduce the risk of type 2 diabetes in mothers with recent gestational diabetes mellitus (GDM). This study compared the cessation of predominant breastfeeding in mothers with and without recent GDM in a multi-ethnic population. METHODS: From May 2008 to May 2010, healthy pregnant women attending antenatal care provided by community health services in Eastern Oslo, Norway were recruited. We included 616 women-58% non-Western-and interviewed and examined them at a mean of 15 and 28 weeks of gestation and 14 weeks' postpartum. Cox regression models examined the association between GDM, as assessed by the 2013 World Health Organization criteria, and breastfeeding cessation. RESULTS: Overall, 190 of the 616 (31%) mothers had GDM and they ended predominant breastfeeding earlier than mothers without GDM, with an adjusted hazard ratio (aHR) of 1.33 and 95% confidence interval (95% CI) of 1.01-1.77. Mothers of South Asian origin ended predominant breastfeeding earlier than Western European mothers in the adjusted analysis (aHR 1.53, 95% CI: 1.04-2.25), but Middle Eastern mothers did not. CONCLUSION: Recent gestational diabetes was associated with earlier cessation of predominant breastfeeding in Western European and non-Western women.

A Key Characteristic of Sex Differences in the Developing Brain: Greater Variability in Brain Structure of Boys than Girls.

In many domains, including cognition and personality, greater variability is observed in males than in females in humans. However, little is known about how variability differences between sexes are represented in the brain. The present study tested whether there is a sex difference in variance in brain structure using a cohort of 643 males and 591 females aged between 3 and 21 years. The broad age-range of the sample allowed us to test if variance differences in the brain differ across age. We observed significantly greater male than female variance for several key brain structures, including cerebral white matter and cortex, hippocampus, pallidum, putamen, and cerebellar cortex volumes. The differences were observed at both upper and lower extremities of the distributions and appeared stable across development. These findings move beyond mean levels by showing that sex differences were pronounced for variability, thereby providing a novel perspective on sex differences in the developing brain.

Explaining socioeconomic inequalities in exclusive breast feeding in Norway.

BACKGROUND AND OBJECTIVE: In high-income countries, lower socioeconomic position is associated with lower rates of breast feeding, but it is unclear what factors explain this inequality. Our objective was to examine the association between socioeconomic position and exclusive breast feeding, and to explore whether socioeconomic inequality in exclusive breast feeding could be explained by other sociodemographic characteristics, for example, maternal age and parity, smoking habits, birth characteristics, quality of counselling and breastfeeding difficulties. METHODS: We used data from a questionnaire sent to mothers when their infants were five completed months as part of a trial of a breastfeeding intervention in Norway. We used maternal education as an indicator of socioeconomic position. Analyses of 1598 mother-infant pairs were conducted using logistic regression to assess explanatory factors of educational inequalities in breast feeding. RESULTS: Socioeconomic inequalities in exclusive breast feeding were present from the beginning and persisted for five completed months, when 22% of the most educated mothers exclusively breast fed compared with 7% of the least educated mothers: OR 3.39 (95% CI 1.74 to 6.61). After adjustment for all potentially explanatory factors, the OR was reduced to 1.49 (95% CI 0.70 to 3.14). This decrease in educational inequality seemed to be mainly driven by sociodemographic factors, smoking habits and breastfeeding difficulties, in particular perceived milk insufficiency. CONCLUSIONS: Socioeconomic inequalities in exclusive breast feeding at 5 months were largely explained by sociodemographic factors, but also by modifiable factors, such as smoking habits and breastfeeding difficulties, which can be amenable to public health interventions. TRIAL REGISTRATION NUMBER: NCT01025362.

Tomato-based randomized controlled trial in prostate cancer patients: Effect on PSA.

BACKGROUND & AIMS: The effect of lycopene-containing foods in prostate cancer development remains undetermined. We tested whether a lycopene-rich tomato intervention could reduce the levels of prostate specific antigen (PSA) in prostate cancer patients. METHODS: Prior to their curative treatment, 79 patients with prostate cancer were randomized to a nutritional intervention with either 1) tomato products containing 30 mg lycopene per day; 2) tomato products plus selenium, omega-3 fatty acids, soy isoflavones, grape/pomegranate juice, and green/black tea (tomato-plus); or 3) control diet for 3 weeks. RESULTS: The main analysis, which included patients in all risk categories, did not reveal differences in changes of PSA-values between the intervention and control groups. Post-hoc, exploratory analyses within intermediate risk (n = 41) patients based on tumor classification and Gleason score post-surgery, revealed that median PSA decreased significantly in the tomato group as compared to controls (-2.9% and +6.5% respectively, p = 0.016). In separate post-hoc analyses, we observed that median PSA-values decreased by 1% in patients with the highest increases in plasma lycopene, selenium and C20:5 n-3 fatty acid, compared to an 8.5% increase in the patients with the lowest increase in lycopene, selenium and C20:5 n-3 fatty acid (p = 0.003). Also, PSA decreased in patients with the highest increase in lycopene alone (p = 0.009). CONCLUSIONS: Three week nutritional interventions with tomato-products alone or in combination with selenium and n-3 fatty acids lower PSA in patients with non-metastatic prostate cancer. Our observation suggests that the effect may depend on both aggressiveness of the disease and the blood levels of lycopene, selenium and omega-3 fatty acids.

Sensitivity of regression calibration to non-perfect validation data with application to the Norwegian Women and Cancer Study.

Measurement error occurs when we observe error-prone surrogates, rather than true values. It is common in observational studies and especially so in epidemiology, in nutritional epidemiology in particular. Correcting for measurement error has become common, and regression calibration is the most popular way to account for measurement error in continuous covariates. We consider its use in the context where there are validation data, which are used to calibrate the true values given the observed covariates. We allow for the case that the true value itself may not be observed in the validation data, but instead, a so-called reference measure is observed. The regression calibration method relies on certain assumptions.This paper examines possible biases in regression calibration estimators when some of these assumptions are violated. More specifically, we allow for the fact that (i) the reference measure may not necessarily be an 'alloyed gold standard' (i.e., unbiased) for the true value; (ii) there may be correlated random subject effects contributing to the surrogate and reference measures in the validation data; and (iii) the calibration model itself may not be the same in the validation study as in the main study; that is, it is not transportable. We expand on previous work to provide a general result, which characterizes potential bias in the regression calibration estimators as a result of any combination of the violations aforementioned. We then illustrate some of the general results with data from the Norwegian Women and Cancer Study.