Nynke Smidt

Evidence of bias and variation in diagnostic accuracy studies

Background: Studies with methodologic shortcomings can overestimate the accuracy of a medical test. We sought to determine and compare the direction and magnitude of the effects of a number of potential sources of bias and variation in studies on estimates of diagnostic accuracy. Methods: We identified meta-analyses of the diagnostic accuracy of tests through an electronic search of the databases MEDLINE, EMBASE, DARE and MEDION (1999–2002). We included meta-analyses with at least 10 primary studies without preselection based on design features. Pairs of reviewers independently extracted study characteristics and original data from the primary studies. We used a multivariable meta-epidemiologic regression model to investigate the direction and strength of the association between 15 study features on estimates of diagnostic accuracy. Results: We selected 31 meta-analyses with 487 primary studies of test evaluations. Only 1 study had no design deficiencies. The quality of reporting was poor in most of the studies. We found significantly higher estimates of diagnostic accuracy in studies with nonconsecutive inclusion of patients (relative diagnostic odds ratio [RDOR] 1.5, 95% confidence interval [CI] 1.0–2.1) and retrospective data collection (RDOR 1.6, 95% CI 1.1–2.2). The estimates were highest in studies that had severe cases and healthy controls (RDOR 4.9, 95% CI 0.6–37.3). Studies that selected patients based on whether they had been referred for the index test, rather than on clinical symptoms, produced significantly lower estimates of diagnostic accuracy (RDOR 0.5, 95% CI 0.3–0.9). The variance between meta-analyses of the effect of design features was large to moderate for type of design (cohort v. case–control), the use of composite reference standards and the use of differential verification; the variance was close to zero for the other design features. Interpretation: Shortcomings in study design can affect estimates of diagnostic accuracy, but the magnitude of the effect may vary from one situation to another. Design features and clinical characteristics of patient groups should be carefully considered by researchers when designing new studies and by readers when appraising the results of such studies. Unfortunately, incomplete reporting hampers the evaluation of potential sources of bias in diagnostic accuracy studies.

Corticosteroid injections for lateral epicondylitis: a systematic review

&NA; Patients with lateral epicondylitis (tennis elbow) are frequently treated with corticosteroid injections, in order to relieve pain and diminish disability. The objective of this review was to evaluate the effectiveness of corticosteroid injections for lateral epicondylitis. Randomised controlled trials (RCTs) were identified by a highly sensitive search strategy in six databases in combination with reference tracking. Two independent reviewers selected and assessed the methodological quality of RCTs that included patients with lateral epicondylitis treated with corticosteroid injection(s), and reported at least one clinically relevant outcome measure. Standardised mean differences were computed for continuous data and relative risks (RR) for dichotomous data. A best‐evidence synthesis was conducted, weighting the studies with respect to their internal validity, statistical significance, clinical relevance, and statistical power. Thirteen studies consisting of 15 comparisons were included in the review, evaluating the effects of corticosteroid injections compared to placebo injection (n=2), injection with local anaesthetic (n=5), another conservative treatment (n=5), or another corticosteroid injection (n=3). Almost all studies had poor internal validity scores. For short‐term outcomes (≤6 weeks), statistically significant and clinically relevant differences were found on pain, global improvement and grip strength for corticosteroid injection compared to placebo, local anaesthetic and conservative treatments. For intermediate (6 weeks–6 months) and long‐term outcomes (≥6 months), no statistically significant or clinically relevant results in favour of corticosteroid injections were found. Although the available evidence shows superior short‐term effects of corticosteroid injections for lateral epicondylitis, it is not possible to draw firm conclusions on the effectiveness of injections, due to the lack of high quality studies. No beneficial effects were found for intermediate or long‐term follow‐up. More, better designed, conducted and reported RCTs with intermediate and long‐term follow‐up are needed.

Cohort Profile: LifeLines, a three-generation cohort study and biobank

The LifeLines Cohort Study is a large population-based cohort study and biobank that was established as a resource for research on complex interactions between environmental, phenotypic and genomic factors in the development of chronic diseases and healthy ageing. Between 2006 and 2013, inhabitants of the northern part of The Netherlands and their families were invited to participate, thereby contributing to a three-generation design. Participants visited one of the LifeLines research sites for a physical examination, including lung function, ECG and cognition tests, and completed extensive questionnaires. Baseline data were collected for 167 729 participants, aged from 6 months to 93 years. Follow-up visits are scheduled every 5 years, and in between participants receive follow-up questionnaires. Linkage is being established with medical registries and environmental data. LifeLines contains information on biochemistry, medical history, psychosocial characteristics, lifestyle and more. Genomic data are available including genome-wide genetic data of 15 638 participants. Fasting blood and 24-h urine samples are processed on the day of collection and stored at -80 °C in a fully automated storage facility. The aim of LifeLines is to be a resource for the national and international scientific community. Requests for data and biomaterials can be submitted to the LifeLines Research Office [LLscience@umcg.nl].

Systematic review: accuracy of anti-citrullinated Peptide antibodies for diagnosing rheumatoid arthritis

BACKGROUND Early recognition and treatment of rheumatoid arthritis is important to prevent irreversible joint damage. Anti-citrullinated peptide antibodies (ACPA) have been suggested for early diagnosis. PURPOSE To compare the accuracy of ACPA and rheumatoid factor in diagnosing rheumatoid arthritis in patients with early symptoms of the disease. DATA SOURCES 10 medical databases from inception to September 2009, with no language or publication restrictions, and references of included studies. STUDY SELECTION Two independent reviewers screened searches. Full articles were assessed by one reviewer and checked by a second reviewer to identify studies that reported 2 x 2 data on ACPA for the diagnosis of rheumatoid arthritis (by 1987 American College of Rheumatology criteria). DATA EXTRACTION One reviewer abstracted data on patient characteristics, ACPA details, and 2 x 2 data and assessed study quality by using the QUADAS tool. A second reviewer checked extractions. DATA SYNTHESIS 151 studies were included, with considerable heterogeneity in sensitivity (range, 12% to 93%) and specificity (range, 63% to 100%). In cohort studies that investigated second-generation anti-cyclic citrullinated peptide antibodies (anti-CCP2) in patients with early rheumatoid arthritis (<2 years), summary sensitivity and specificity were 57% (95% CI, 51% to 63%) and 96% (CI, 93% to 97%), respectively. Case-control and cross-sectional studies and studies of patients with established rheumatoid arthritis all overestimated sensitivity. Anti-CCP2 had greater specificity than rheumatoid factor (96% vs. 86%), with similar sensitivity. Evidence was insufficient to ascertain whether the combination of anti-CCP2 and rheumatoid factor provides additional benefit over anti-CCP2 alone. LIMITATIONS Most studies used a diagnostic case-control design, which overestimated sensitivity. Items relating to study quality were rarely reported. Publication bias could not be assessed. CONCLUSION Anti-CCP2 should be included in the work-up of patients with early symptoms of rheumatoid arthritis.

Social relationships and risk of dementia: A systematic review and meta-analysis of longitudinal cohort studies

It is unclear to what extent poor social relationships are related to the development of dementia. A comprehensive systematic literature search identified 19 longitudinal cohort studies investigating the association between various social relationship factors and incident dementia in the general population. Relative risks (RRs) with 95% confidence intervals (CIs) were pooled using random-effects meta-analysis. Low social participation (RR: 1.41 (95% CI: 1.13-1.75)), less frequent social contact (RR: 1.57 (95% CI: 1.32-1.85)), and more loneliness (RR: 1.58 (95% CI: 1.19-2.09)) were statistically significant associated with incident dementia. The results of the association between social network size and dementia were inconsistent. No statistically significant association was found for low satisfaction with social network and the onset of dementia (RR: 1.25 (95% CI: 0.96-1.62). We conclude that social relationship factors that represent a lack of social interaction are associated with incident dementia. The strength of the associations between poor social interaction and incident dementia is comparable with other well-established risk factors for dementia, including low education attainment, physical inactivity, and late-life depression.

Representativeness of the LifeLines Cohort Study

Background LifeLines is a large prospective population-based three generation cohort study in the north of the Netherlands. Different recruitment strategies were adopted: recruitment of an index population via general practitioners, subsequent inclusion of their family members, and online self-registration. Our aim was to investigate the representativeness of the adult study population at baseline and to evaluate differences in the study population according to recruitment strategy. Methods Demographic characteristics of the LifeLines study population, recruited between 2006–2013, were compared with the total adult population in the north of the Netherlands as registered in the Dutch population register. Socioeconomic characteristics, lifestyle, chronic diseases, and general health were further compared with participants of the Permanent Survey of Living Conditions within the region (2005–2011, N = 6,093). Differences according to recruitment strategy were assessed. Results Compared with the population of the north of the Netherlands, LifeLines participants were more often female, middle aged, married, living in a semi-urban place and Dutch native. Adjusted for differences in demographic composition, in LifeLines a smaller proportion had a low educational attainment (5% versus 14%) or had ever smoked (54% versus 66%). Differences in the prevalence of various chronic diseases and low general health scores were mostly smaller than 3%. The age profiles of the three recruitment groups differed due to age related inclusion criteria of the recruitment groups. Other differences according to recruitment strategy were small. Conclusions Our results suggest that, adjusted for differences in demographic composition, the LifeLines adult study population is broadly representative for the adult population of the north of the Netherlands. The recruitment strategy had a minor effect on the level of representativeness. These findings indicate that the risk of selection bias is low and that risk estimates in LifeLines can be generalized to the general population.

Reporting Diagnostic Accuracy Studies: Some Improvements after 10 Years of STARD

PURPOSE To evaluate how diagnostic accuracy study reports published in 2012 adhered to the Standards for Reporting of Diagnostic Accuracy (STARD) statement and whether there were any differences in reporting compared with 2000 and 2004. MATERIALS AND METHODS PubMed was searched for studies published in 12 high-impact-factor journals in 2012 that evaluated the accuracy of one or more diagnostic tests against a clinical reference standard. Two independent reviewers scored reporting completeness of each article with the 25-item STARD checklist. Mixed-effects modeling was used to analyze differences in reporting with previous evaluations from articles published in 2000 and 2004. RESULTS Included were 112 articles. The overall mean number of STARD items reported in 2012 was 15.3 ± 3.9 (standard deviation; range, 6.0-23.5). There was an improvement of 3.4 items (95% confidence interval: 2.6, 4.3) compared with studies published in 2000, and an improvement of 1.7 items (95% confidence interval: 0.9, 2.5) compared with studies published in 2004. Significantly more items were reported for single-gate studies compared with multiple-gate studies (16.8 vs 12.1, respectively; P < .001) and for studies that evaluated imaging tests compared with laboratory tests and other types of tests (17.0 vs 14.0 vs 14.5, respectively; P < .001). CONCLUSION Completeness of reporting improved in the 10 years after the launch of STARD, but it remains suboptimal for many articles. Reporting of inclusion criteria and sampling methods for recruiting patients, information about blinding, and confidence intervals for accuracy estimates are in need of further improvement.

Cochrane corner: is integrated disease management for patients with COPD effective?

Patients with COPD experience respiratory symptoms, impairments of daily living and recurrent exacerbations. The aim of integrated disease management (IDM) is to establish a programme of different components of care (ie, self-management, exercise, nutrition) in which several healthcare providers (ie, nurses, general practitioners, physiotherapists, pulmonologists) collaborate to provide efficient and good quality of care. The aim of this Cochrane systematic review was to evaluate the effectiveness of IDM on quality of life, exercise tolerance and exacerbation related outcomes. Searches for all available evidence were carried out in various databases. Included randomised controlled trials (RCTs) consisted of interventions with multidisciplinary (≥2 healthcare providers) and multitreatment (≥2 components) IDM interventions with duration of at least 3 months. Two reviewers independently searched, assessed and extracted data of all RCTs. A total of 26 RCTs were included, involving 2997 patients from 11 different countries with a follow-up varying from 3 to 24 months. In all 68% of the patients were men, with a mean age of 68 years and a mean forced expiratory volume in 1 s (FEV1) predicted value of 44.3%. Patients treated with an IDM programme improved significantly on quality of life scores and reported a clinically relevant improvement of 44 m on 6 min walking distance, compared to controls. Furthermore, the number of patients with ≥1 respiratory related hospital admission reduced from 27 to 20 per 100 patients. Duration of hospitalisation decreased significantly by nearly 4 days.

Subjective burden among spousal and adult-child informal caregivers of older adults: results from a longitudinal cohort study

BackgroundPressures on informal caregivers are likely to increase due to increasing life expectancy and health care costs, which stresses the importance of prevention of subjective burden. The present study examined the correlates of overall subjective burden and multiple burden dimensions among spousal and adult-child caregivers of Dutch older adults, both cross-sectional and longitudinal (12-months follow-up).MethodsIn 2010 and 2011 baseline and follow-up data was collected in a sample of informal caregivers and care recipients in the Northern provinces of the Netherlands. Subjective burden included 7 burden dimensions and a summary score for overall subjective burden, based on the Care-Related Quality of Life Instrument (CarerQoL-7D). Objective stressors were the time investment in caregiving (hours of household care, personal care, practical care) and the health situation of the care recipient, including multimorbidity, functional limitations (Katz Index of Independence Basic Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL), and cognitive functioning problems (EQ-5D + C). Correlates of subjective burden were evaluated with linear and logistic regression analyses.ResultsThe sample consisted of 356 caregivers at baseline (43% spousal, 57% adult-child caregivers), and 158 caregivers at follow-up (45% spousal, 55% adult-child caregivers). At baseline and follow-up, spousal caregivers experienced a higher overall subjective burden, and reported more often mental health problems, physical health problems, and problems with combining daily activities, compared to adult-child caregivers. For spousal caregivers, a poorer health situation of the care recipient was associated with higher subjective burden, while adult-child caregivers reported higher levels of subjective burden when their time investment in caregiving was high. Subjective burden at follow-up was mainly explained by baseline subjective burden.ConclusionsThese results indicate that for effective caregiver support, it is crucial to take the type of care relationship into account, since the level and correlates of overall subjective burden and burden dimensions varied for spousal and adult-child caregivers. In addition, reducing subjective burden will also positively impact the subjective burden over time.

SORTA: a system for ontology-based re-coding and technical annotation of biomedical phenotype data.

There is an urgent need to standardize the semantics of biomedical data values, such as phenotypes, to enable comparative and integrative analyses. However, it is unlikely that all studies will use the same data collection protocols. As a result, retrospective standardization is often required, which involves matching of original (unstructured or locally coded) data to widely used coding or ontology systems such as SNOMED CT (clinical terms), ICD-10 (International Classification of Disease) and HPO (Human Phenotype Ontology). This data curation process is usually a time-consuming process performed by a human expert. To help mechanize this process, we have developed SORTA, a computer-aided system for rapidly encoding free text or locally coded values to a formal coding system or ontology. SORTA matches original data values (uploaded in semicolon delimited format) to a target coding system (uploaded in Excel spreadsheet, OWL ontology web language or OBO open biomedical ontologies format). It then semi- automatically shortlists candidate codes for each data value using Lucene and n-gram based matching algorithms, and can also learn from matches chosen by human experts. We evaluated SORTA’s applicability in two use cases. For the LifeLines biobank, we used SORTA to recode 90 000 free text values (including 5211 unique values) about physical exercise to MET (Metabolic Equivalent of Task) codes. For the CINEAS clinical symptom coding system, we used SORTA to map to HPO, enriching HPO when necessary (315 terms matched so far). Out of the shortlists at rank 1, we found a precision/recall of 0.97/0.98 in LifeLines and of 0.58/0.45 in CINEAS. More importantly, users found the tool both a major time saver and a quality improvement because SORTA reduced the chances of human mistakes. Thus, SORTA can dramatically ease data (re)coding tasks and we believe it will prove useful for many more projects. Database URL: http://molgenis.org/sorta or as an open source download from http://www.molgenis.org/wiki/SORTA