Detailed Patient-level Information Research Articles

A unified approach for synthesizing population-level covariate effect information in semiparametric estimation with survival data.

There has been a growing interest in developing methodologies to combine information from public domains to improve efficiency in the analysis of relatively small-scale studies that collect more detailed patient-level information. The auxiliary information is usually given in the form of summary statistics or regression coefficients. Thus, the question arises as to how to incorporate the summary information in the model estimation procedure. In this article, we consider statistical analysis of right-censored survival data when additional information about the covariate effects evaluated in a reduced Cox model is available. Recognizing that such external information can be summarized using population moments, we present a unified framework by employing the generalized method of moments to combine information from different sources for the analysis of survival data. The proposed estimator can be shown to be consistent and asymptotically normal; moreover, it is more efficient than the maximum partial likelihood estimator. We also consider incorporating uncertainty of the external information in the inference procedure. Simulation studies show that, by incorporating the additional summary information, the proposed estimators enjoy a substantial gain in efficiency over the conventional approach. A data analysis of a pancreatic cancer cohort study is presented to illustrate the methods and theory.

Bridging of Neisseria Gonorrhoeae Across Diverse Sexual Networks in the HIV Pre-Exposure Prophylaxis (PrEP) Era: A Clinical and Molecular Epidemiological Study

Background: Despite 'best effort' public health measures, the incidence of gonorrhoea is increasing in many countries. Recently, whole genome sequencing (WGS) has been used to investigate transmission of Neisseria gonorrhoeae, including antimicrobial-resistant (AMR) lineages, but to date, most studies have not combined genomic data with detailed patient-level information on sexual behaviour to define the extent of transmission across population subgroups ('bridging'). Methods: We performed an observational study and undertook WGS and bioinformatic analysis on all cultured clinical isolates of N. gonorrhoeae in the state of Victoria, Australia, from January to December 2017. Antimicrobial susceptibility testing was performed on all isolates, and detailed epidemiological data was obtained, including data on sexual orientation, HIV status, use of HIV pre-exposure prophylaxis (PrEP), sex work, and overseas travel. Epidemiological associations were made with dominant genetic clusters of N. gonorrhoeae. Findings: A total of 2,186 isolates were sequenced from 2,055 patients, 86·3% of whom were male. We identified eleven dominant genetic clusters, containing thirty or more patients, with the largest cluster comprising 181 patients. There was extensive bridging of clusters between men who have sex with men (MSM) and heterosexuals, with bisexual males identified in seven of the major clusters. Eight major clusters contained HIV-positive and HIV-negative patients (including individuals receiving PrEP). We also identified transmission of a novel azithromycin-resistant clone, associated with a mosaic mtr locus. Interpretation: To our knowledge, our study is the first to combine WGS with comprehensive individual-level behavioural risk data, providing verification for transmission of multiple gonococcal lineages within and across distinct sexual networks. Application of these methods in real-time will allow gonorrhoea transmission and antimicrobial resistance to be tracked, with 'hotspots' identified for interventions aimed at improving gonorrhoea control. Funding Statement: The National Health and Medical Research Council of Australia, and the Victorian Department of Health and Human Services. DAW (GNT1123854), EPFC (GNT1091226), and JCK (GNT1142613) are supported by Early Career Fellowships from the National Health and Medical Research Council (NHMRC) of Australia. BPH is supported by a Practitioner Fellowship from the NHMRC (GNT1105905). DJI is supported by the European Union’s Horizon 2020 research and innovation programme under grant agreement 643476. Work in this study was supported by a Project Grant from the NHMRC (GNT1147735) and a Partnership grant from the NHMRC (GNT1149991). MDU PHL is funded by the Victorian Department of Health and Human Services. Declaration of Interests: The authors declare no competing interests. Ethics Approval Statement: Data were collected in accordance with the Victorian Public Health and Wellbeing Act 2008. Ethical approval was obtained from the Alfred Hospital Ethics Committee (Project 625/17).

Multi-method laboratory user evaluation of an actionable clinical performance information system: Implications for usability and patient safety

IntroductionElectronic audit and feedback (e-A&F) systems are used worldwide for care quality improvement. They measure health professionals’ performance against clinical guidelines, and some systems suggest improvement actions. However, little is known about optimal interface designs for e-A&F, in particular how to present suggested actions for improvement. We developed a novel theory-informed system for primary care (the Performance Improvement plaN GeneratoR; PINGR) that covers the four principal interface components: clinical performance summaries; patient lists; detailed patient-level information; and suggested actions. As far as we are aware, this is the first report of an e-A&F system with all four interface components. Objectives(1) Use a combination of quantitative and qualitative methods to evaluate the usability of PINGR with target end-users; (2) refine existing design recommendations for e-A&F systems; (3) determine the implications of these recommendations for patient safety. MethodsWe recruited seven primary care physicians to perform seven tasks with PINGR, during which we measured on-screen behaviour and eye movements. Participants subsequently completed usability questionnaires, and were interviewed in-depth. Data were integrated to: gain a more complete understanding of usability issues; enhance and explain each other’s findings; and triangulate results to increase validity. ResultsParticipants committed a median of 10 errors (range 8–21) when using PINGR’s interface, and completed a median of five out of seven tasks (range 4–7). Errors violated six usability heuristics: clear response options; perceptual grouping and data relationships; representational formats; unambiguous description; visually distinct screens for confusable items; and workflow integration. Eye movement analysis revealed the integration of components largely supported effective user workflow, although the modular design of clinical performance summaries unnecessarily increased cognitive load. Interviews and questionnaires revealed PINGR is user-friendly, and that improved information prioritisation could further promote useful user action. ConclusionsComparing our results with the wider usability literature we refine a previously published set of interface design recommendations for e-A&F. The implications for patient safety are significant regarding: user engagement; actionability; and information prioritisation. Our results also support adopting multi-method approaches in usability studies to maximise issue discovery and the credibility of findings.

Physical frailty and functional status in patients with advanced kidney disease: a protocol for a systematic review

BackgroundPredicting outcomes in those with chronic kidney disease or following transplantation is challenging, and current models lack detailed patient-level information. Frailty and poor functional status are risk factors for adverse patient outcomes that may be useful additions to prognostic tools in patients with chronic kidney disease. The purpose of this systematic review is to examine whether frailty or functional status are associated with increased risk of mortality or adverse clinical outcomes in patients with advanced kidney disease.Methods/designWe will conduct a systematic review to identify and evaluate studies linking frailty and functional status with patient outcomes in populations with advanced kidney disease. We will search MEDLINE, Embase, and the Cochrane Central Register for Controlled Trials. Two reviewers will conduct all screening and data extraction independently. A modified version of the Quality In Prognosis Studies tool will be used to evaluate the quality of the studies. If meta-analysis of outcome data is possible, a random effects model will be used.DiscussionThe results of this review will inform the development, selection, and validation of appropriate metrics needed to improve prognostication in patients with chronic kidney disease.Systematic review registrationPROSPERO CRD42016045251

Gestational Age of Delivery in Pregnancies Complicated by Chronic Hypertension.

To identify the gestational age of planned delivery in pregnancies complicated by chronic hypertension that minimizes the risk of perinatal death and severe adverse events. This was a retrospective cohort study of all singletons complicated by hypertension. Detailed patient-level information was collected by chart review, including indication for delivery. Planned delivery at 36-36 6/7, 37-37 6/7, 38-38 6/7, and 39-39 6/7 weeks of gestation was compared with expectant management beyond each respective gestational age. Patients were excluded for fetal anomalies, inaccurate dating, and major medical problems other than hypertension, diabetes, or renal disease. The primary outcome was a composite of stillbirth, neonatal death, assisted ventilation, cord pH less than 7.0, 5-minute Apgar score of 3 or less, and neonatal seizures. Secondary outcomes were preeclampsia, severe preeclampsia, primary cesarean delivery, and neonatal length of stay greater than 5 days. Groups were compared using Student's t test and χ tests. Six hundred eighty-three women with hypertension reached 36 weeks of gestation. Patients with planned delivery at less than 39 weeks of gestation were more likely to have baseline renal disease. Before 37 weeks of gestation, planned delivery was associated with a statistically significant increase in the primary composite adverse neonatal outcome (10.0% compared with 2.6%, P=.04); after 38 weeks of gestation, expectant management was associated with a nonstatistically significant increase in the primary composite outcome (0% compared with 2.3%, P=.40). Expectant management beyond 39 weeks of gestation was associated with a statistically significant increase in severe preeclampsia (0% compared with 10.3%, P=.001). Expectant management beyond 39 weeks of gestation was associated with increasing incidence of severe preeclampsia; planned delivery before 37 weeks of gestation was associated with an increase in adverse neonatal outcomes. Further well-powered studies are needed to delineate the optimal gestational age of delivery.

Confounding Adjustment in Comparative Effectiveness Research Conducted Within Distributed Research Networks

A distributed research network (DRN) of electronic health care databases, in which data reside behind the firewall of each data partner, can support a wide range of comparative effectiveness research (CER) activities. An essential component of a fully functional DRN is the capability to perform robust statistical analyses to produce valid, actionable evidence without compromising patient privacy, data security, or proprietary interests. We describe the strengths and limitations of different confounding adjustment approaches that can be considered in observational CER studies conducted within DRNs, and the theoretical and practical issues to consider when selecting among them in various study settings. Several methods can be used to adjust for multiple confounders simultaneously, either as individual covariates or as confounder summary scores (eg, propensity scores and disease risk scores), including: (1) centralized analysis of patient-level data, (2) case-centered logistic regression of risk set data, (3) stratified or matched analysis of aggregated data, (4) distributed regression analysis, and (5) meta-analysis of site-specific effect estimates. These methods require different granularities of information be shared across sites and afford investigators different levels of analytic flexibility. DRNs are growing in use and sharing of highly detailed patient-level information is not always feasible in DRNs. Methods that incorporate confounder summary scores allow investigators to adjust for a large number of confounding factors without the need to transfer potentially identifiable information in DRNs. They have the potential to let investigators perform many analyses traditionally conducted through a centralized dataset with detailed patient-level information.

Race/Ethnicity, Serum 25-Hydroxyvitamin D, and Heart Disease

A growing body of observational evidence had suggested that low 25-hydroxyvitamin D (25[OH]D) concentrationsmay be a key biological predictor of increased rates of coronary heart disease (CHD).1Because lowserum25(OH)Dconcentrationsare more common and more severe among racial/ethnic minority groups,which are also affected by higher rates of CHD and CHD risk factors, low vitamin D was being heralded as a potentialmodifiablecontributor to racial/ethnicdisparities incardiovascular health.2 However, the causal link between 25 (OH)D concentrations and CHD remains uncertain.3,4 Additionally, the potential clinical implications of low serum 25(OH)Dconcentrations, thepathologicalmechanismsthrough which vitamin Dmaymodulate CHD, and whether these factors differ across racial/ethnic groups are unclear. Todate, the literatureonvitaminDandCHDhasbeen limited due to larger studies being observational in nature. Randomized interventions are smaller, shorter in duration, and only addressCHDsurrogates. Previous studies have also been inconsistent in defining vitamin D status, in defining and determining cardiovascular outcomes, in the use of baseline 25 (OH)Dconcentrations, and in adjusting for seasonal variations in vitamin D concentrations.1 Further, in an attempt to address racial/ethnic influences on possible vitamin D–related health outcomes, the literature is limited by the lack of detailed patient-level information and adequate numbers of diverse populations. In this issue of JAMA, Robinson-Cohen and colleagues5 conducted a well-designed investigation to assess the associationof serum25(OH)Dconcentrationswith time to first adjudicatedCHDevent acrossmultiple racial groups and to provide information to help unravel the complex associations between race/ethnicity, 25(OH)D, and CHD. The authors analyzed data from 6436 diverse racial/ethnic study participants in a community-based prospective cohort, the Multi-Ethnic Study of Atherosclerosis (MESA). From 2000 to 2002, participants between 45 and 84 years of age without known cardiovascular disease at baseline were enrolled from 6 sites. Detaileddemographic, clinical, and laboratorydatawereobtained and for the present analysis participants had a mean follow-upof8.5 years. Theparticipants included38%white, 28% black, 22% Hispanic, and 12% Chinese individuals who completed detailed interviews including estimated vitamin D intake to complement themeasurement of 25(OH)D concentrations and in-person examinations assessing CHD risk factors. During the follow-up period, 361 study participants developed an incident CHD event (including myocardial infarction, angina, or CHD death). Robinson-Cohenetal estimated the relativehazardofCHD using 3 different Cox proportional hazards regression models. The first model included demographic data; the second model added likely potential confounding variables such as bodymass index, smokingstatus,educationalattainment, family income,physicalactivity,andvitaminDintake;andthethird model addedpossibly confounding ormediating variables including diabetes status, systolic blood pressure, chronic kidney disease, antihypertensive and lipid-lowering medications, cholesterol andtriglyceride levels, andC-reactiveprotein concentrations. In the fully adjusted third model, each 10ng/mL decrement in serum 25(OH)D concentrations was associatedwith a significant 15% greater overall risk of incident CHD. However, the investigators noted substantial differencesamongracial/ethnicgroups regarding theassociationbetween 25(OH)D concentrations and CHD, with a 26% greater risk of incident CHD among white participants and a 76% greater risk amongChinese participants. Unexpectedly, there was no significant association between 25(OH)D concentrations and CHD risk among black or Hispanic participants. In this study the lowestmeanbaselineserum25(OH)Dconcentrationswere found inblackparticipants (19.2ng/mL),moderately low 25(OH)D concentrations in Hispanic (24.6 ng/mL) andChineseparticipants (26.7ng/mL),andadequate levels (≥30 ng/mL) inwhiteparticipants (30.1ng/mL).Basedonthesebaseline findings, itmight be anticipated that if the serumconcentrationof 25(OH)Dwas a strongpredictor of CHD, the relationshipwouldbemostpronounced inblackparticipantswhohad the lowest 25(OH)Dconcentrationsatbaseline.However, these findings did not support the anticipated relationship between low 25(OH)D and increased CHD across all racial/ ethnic groups. Prior studies in nonwhite populations have shown variable findings.4 In a recent 3-month, double-blind, randomized trial of nearly 300 black patients, each 1-ng/mL increase inplasma25-hydroxyvitaminD led to a significant0.2-mmHg reduction in systolic bloodpressure (P = .02), but the studydid not assess CHD events.6 By contrast a double-blind, placebocontrolled study of 90 racial/ethnically diverse patients (ethnicity: 62% Hispanic; race: white 31%, black 18%, Asian 24%, other 27%) with CHD and vitamin D deficiency (<20 ng/mL) found no difference in surrogate markers of cardiovascular health (such as bloodpressure, endothelial function, circulating adhesion molecules, or serum proinflammatory cytokines) amongparticipants receiving 50 000 IUof oral ergocalRelated article page 179 Opinion

Managing and exploiting routinely collected NHS data for research

Health research using routinely collected National Health Service (NHS) data derived from electronic health records (EHRs) and health service information systems has been growing in both importance and quantity. Wide population coverage and detailed patient-level information allow this data to be applied to a variety of research questions. However, the sensitivity, complexity and scale of such data also hamper researchers from fully exploiting this potential. Here, we establish the current challenges preventing researchers from making optimal use of the data sets at their disposal, on both the legislative and practical levels, and give recommendations as to how these challenges can be overcome. A number of projects has recently been launched in the UK to address poor research data-management practices. Rapid Organisation of Healthcare Research Data (ROHRD) at Imperial College, London produced a useful prototype that provides local researchers with a one-stop index of available data sets together with relevant metadata. Increased transparency of data sets' availability and their provenance leads to better utilisation and facilitates compliance with regulatory requirements. Research data resulting from NHS data is often not utilised fully, or is handled in a haphazard manner that prevents full auditability of the research. Furthermore, lack of informatics and data management skills in research teams act as a barrier to implementing more advanced practices, such as provenance capture and detailed, regularly updated, data management strategies. Only by a concerted effort at the levels of research organisations, funding bodies and publishers, can we achieve full transparency and reproducibility of the research.

Data and information requirements for determining veterans' access to health care.

The Department of Veterans Affairs (VA) is responding to changing requirements for decision-support data by maximizing the value of data contained in VA and non-VA sources. The data are used to answer questions relating to the accessibility and utilization of VA and non-VA health services. Access studies require accurate estimates of the number of persons served and the number of persons who could be served. To derive these population estimates, VA employs census data to develop projections of the veteran population at the national, state, and county levels. Data from many surveys are used to supplement the census data. Access studies also require quantitative and qualitative data on the characteristics of VA and non-VA health care delivery systems at the national, state, and local levels. The Department of Veterans Affairs obtains health care system data from external sources, including the US Department of Health and Human Services, the American Medical Association, and the American Hospital Association, and from internal sources, including VA surveys and the VA administration inpatient and outpatient files. Utilization studies need more detailed patient-level information than access studies. Data elements pertaining to the reason for health care encounters and the services rendered are obtained from survey data, the VA inpatient and outpatient administration files, the national Medicare database, and state Medicaid databases. The Department of Veterans Affairs'decision-support analyses for eligibility reform and health care system reform demonstrate the effectiveness of VA in analyzing data from many sources.