Publications

2014
Mandl KD, Kohane IS, McFadden D, Weber GM, Natter M, Mandel J, Schneeweiss S, Weiler S, Klann JG, Bickel J, et al. Scalable Collaborative Infrastructure for a Learning Healthcare System (SCILHS): Architecture. J Am Med Inform AssocJ Am Med Inform AssocJ Am Med Inform Assoc. 2014.Abstract
We describe the architecture of the Patient Centered Outcomes Research Institute (PCORI) funded Scalable Collaborative Infrastructure for a Learning Healthcare System (SCILHS, http://www.SCILHS.org) clinical data research network, which leverages the $48 billion dollar federal investment in health information technology (IT) to enable a queryable semantic data model across 10 health systems covering more than 8 million patients, plugging universally into the point of care, generating evidence and discovery, and thereby enabling clinician and patient participation in research during the patient encounter. Central to the success of SCILHS is development of innovative 'apps' to improve PCOR research methods and capacitate point of care functions such as consent, enrollment, randomization, and outreach for patient-reported outcomes. SCILHS adapts and extends an existing national research network formed on an advanced IT infrastructure built with open source, free, modular components.
Dunn AG, Coiera E, Mandl KD. Is Biblioleaks inevitable?. J Med Internet ResJ Med Internet ResJ Med Internet Res. 2014;16:e112.Abstract
In 2014, the vast majority of published biomedical research is still hidden behind paywalls rather than open access. For more than a decade, similar restrictions over other digitally available content have engendered illegal activity. Music file sharing became rampant in the late 1990s as communities formed around new ways to share. The frequency and scale of cyber-attacks against commercial and government interests has increased dramatically. Massive troves of classified government documents have become public through the actions of a few. Yet we have not seen significant growth in the illegal sharing of peer-reviewed academic articles. Should we truly expect that biomedical publishing is somehow at less risk than other content-generating industries? What of the larger threat--a "Biblioleaks" event--a database breach and public leak of the substantial archives of biomedical literature? As the expectation that all research should be available to everyone becomes the norm for a younger generation of researchers and the broader community, the motivations for such a leak are likely to grow. We explore the feasibility and consequences of a Biblioleaks event for researchers, journals, publishers, and the broader communities of doctors and the patients they serve.
Fine AM, Nizet V, Mandl KD.

Participatory medicine: a home score for streptococcal pharyngitis

. Ann Intern Med. 2014;160:289.
Ong MS, Kohane IS, Cai T, Gorman MP, Mandl KD.

Population-Level Evidence for an Autoimmune Etiology of Epilepsy

. JAMA Neurol. 2014.Abstract
IMPORTANCE Epilepsy is a debilitating condition, often with neither a known etiology nor an effective treatment. Autoimmune mechanisms have been increasingly identified. OBJECTIVE To conduct a population-level study investigating the relationship between epilepsy and several common autoimmune diseases. DESIGN, SETTING, AND PARTICIPANTS A retrospective population-based study using claims from a nationwide employer-provided health insurance plan in the United States. Participants were beneficiaries enrolled between 1999 and 2006 (N = 2 518 034). MAIN OUTCOMES AND MEASURES We examined the relationship between epilepsy and 12 autoimmune diseases: type 1 diabetes mellitus, psoriasis, rheumatoid arthritis, Graves disease, Hashimoto thyroiditis, Crohn disease, ulcerative colitis, systemic lupus erythematosus, antiphospholipid syndrome, Sjogren syndrome, myasthenia gravis, and celiac disease. RESULTS The risk of epilepsy was significantly heightened among patients with autoimmune diseases (odds ratio, 3.8; 95% CI, 3.6-4.0; P < .001) and was especially pronounced in children (5.2; 4.1-6.5; P < .001). Elevated risk was consistently observed across all 12 autoimmune diseases. CONCLUSIONS AND RELEVANCE Epilepsy and autoimmune disease frequently co-occur; patients with either condition should undergo surveillance for the other. The potential role of autoimmunity must be given due consideration in epilepsy so that we are not overlooking a treatable cause.
Ong MS, Umetsu DT, Mandl KD.

Consequences of antibiotics and infections in infancy: bugs, drugs, and wheezing

. Ann Allergy Asthma Immunol. 2014.Abstract
BACKGROUND: The prevalence of asthma has increased alarmingly in the past 2 to 3 decades. Increased antibiotic use in infancy has been suggested to limit exposure to gastrointestinal microbes and to predispose to asthma in later life. OBJECTIVE: To evaluate the association between antibiotic exposure during the first year of life and the development of asthma up to the age of 7 years. METHODS: A retrospective population-based study of a cohort of children enrolled in a nationwide employer-provided health insurance plan from January 1, 1999, through December 31, 2006, in the United States (n = 62,576). We evaluated the association between antibiotic exposure during the first year of life and subsequent development of 3 asthma phenotypes: transient wheezing (began and resolved before 3 years of age), late-onset asthma (began after 3 years of age), and persistent asthma (began before 3 years of age and persisted through 4-7 years of age). RESULTS: Antibiotic use in the first year of life was associated with the development of transient wheezing (odds ratio [OR], 2.0; 95% confidence interval [CI], 1.9-2.2; P < .001) and persistent asthma (OR, 1.6; 95% CI, 1.5-1.7; P < .001). A dose-response effect was observed. When 5 or more antibiotic courses were received, the odds of persistent asthma doubled (OR, 1.9; 95% CI, 1.5-2.6; P < .001). There is no association between antibiotic use and late-onset asthma. CONCLUSION: Antibiotic use in the first year life is associated with an increased risk of early-onset childhood asthma that began before 3 years of age. The apparent effect has a clear dose response. Heightened caution about avoiding unnecessary use of antibiotics in infants is warranted.
Natter MD, Ong MS, Ilowite NT, Mandl KD, Mieszkalski KL, Sandborg CI, Wallace C, Schanberg LE, Investigators CR.

A167: variations in patterns of care across pediatric rheumatic diseases in the childhood arthritis & rheumatology alliance network registry

. Arthritis Rheumatol. 2014;66 Suppl 11:S215-6.Abstract
BACKGROUND/PURPOSE: In 2009, the Childhood Arthritis and Rheumatology Research Alliance (CARRA) established a longitudinal multi-center, multiple disease U.S. national registry (CARRA Registry) for pediatric rheumatology with the intent of providing a new framework to drive observational clinical research and best practices, evidence-based care. Simultaneously, recognizing that widely variable therapeutic approaches hinder the ability to conduct meaningful comparative effectiveness studies and pragmatic trials in pediatric rheumatic diseases, CARRA investigators convened expert groups to formulate new consensus-based treatment plans (CTPs) in 5 major pediatric rheumatic disease areas. As the CTP approaches are adopted, it is important to establish baseline treatment variability across pediatric rheumatic diseases and clinical sites in the CARRA network. Using longitudinal data from the CARRA Registry, we provide a first description of variability of care across the network. METHODS: We examine variations of medication usage across 55 clinical sites in the treatment of 8 rheumatic conditions, including juvenile idiopathic arthritis (JIA), SLE and mixed connective tissue disease (MCTD), juvenile dermatomyositis (JDM), localized scleroderma, systemic sclerosis, juvenile primary fibromyalgia syndrome (JPFS), sarcoidosis, and vasculitis. Management of uveitis in JIA patients was also assessed. Study participants include all CARRA registry subjects enrolled in May 2010 through December 2013. Medications were categorized into 4 major classes: biologics, DMARDs, steroids and NSAIDs. We compare the percentage of patients exposed to each medication class at each; care variations were quantified using dispersion measures of standard deviation and range. A subgroup analysis was conducted to assess care variations among the largest group of subjects with similar characteristics of and low disease activity (JIA subjects with an average active joint count of 0 to 1 averaged over the enrolment period), where treatment were hypothesized to be most similar. RESULTS: 8,869 subjects were included in data analysis. Therapeutic approaches were highly variable for all 8 rheumatic diseases (Table 1, Fig 1). Subgroup analysis for JIA showed persistence of variability (Fig 2). CONCLUSION: We quantify a substantial degree of therapeutic practice variability across sites, persisting across disease-severity-matched cohorts. Although enrollment bias is a significant limitation, the magnitude of the variability for the largest cohort (JIA) and persistence across multiple diseases and subtypes supports a widespread effect. This baseline quantification and methods developed for assessing variability will support ongoing efforts to monitor new consensus treatment protocol-based standardization efforts across the CARRA network.
Bourgeois FT, Olson KL, Ioannidis JP, Mandl KD. Association between pediatric clinical trials and global burden of disease. Pediatrics. 2014;133:78-87.Abstract
BACKGROUND: The allocation of research resources should favor conditions responsible for the greatest disease burden. This is particularly important in pediatric populations, which have been underrepresented in clinical research. Our aim was to measure the association between the focus of pediatric clinical trials and burden of disease and to identify neglected clinical domains. METHODS: We performed a cross-sectional study of clinical trials by using trial records in ClinicalTrials.gov. All trials started in 2006 or after and studying patient-level interventions in pediatric populations were included. Age-specific measures of disease burden were obtained for 21 separate conditions for high-, middle-, and low-income countries. We measured the correlation between number of pediatric clinical trials and disease burden for each condition. RESULTS: Neuropsychiatric conditions and infectious diseases were the most studied conditions globally in terms of number of trials (874 and 847 trials, respectively), while intentional injuries (5 trials) and maternal conditions (4 trials) were the least studied. Clinical trials were only moderately correlated with global disease burden (r = 0.58, P = .006). Correlations were also moderate within each of the country income levels, but lowest in low-income countries (r = .47, P = .03). Globally, the conditions most understudied relative to disease burden were injuries (-260 trials for unintentional injuries and -160 trials for intentional injuries), nutritional deficiencies (-175 trials), and respiratory infections (-171 trials). CONCLUSIONS: Pediatric clinical trial activity is only moderately associated with pediatric burden of disease, and least associated in low-income countries. The mismatch between clinical trials and disease burden identifies key clinical areas for focus and investment.
2013
Weitzman ER, Kelemen S, Quinn M, Eggleston EM, Mandl KD. Participatory surveillance of hypoglycemia and harms in an online social network. JAMA Intern Med. 2013;173:345-51.Abstract
IMPORTANCE: Surveillance systems for elucidating the burden of hypoglycemia are limited. OBJECTIVE: To quantify experiences of hypoglycemia and related harms, members of an international online diabetes social network with insulin-dependent diabetes mellitus were polled through a software application ("app"). Aggregate results were returned to participants through network channels. DESIGN: The study period was from March 2011 through April 2012, during which time retrospective reports about experiences with hypoglycemia and related harms were collected from participants using the app. SETTING: The study was undertaken within the TuDiabetes.org international online diabetes social network. PARTICIPANTS: Eligibility criteria included TuDiabetes membership, age 13 years or older, a self-reported diagnosis of diabetes mellitus, ability to read and write English, and Internet access. Of 2827 app users, 687 (response rate, 24.3%) opted in to the volunteer sample. MAIN OUTCOME MEASURES: Primary outcomes included the following: frequency of "going low" (having a low glucose value in the past 2 weeks) and episodes of severe hypoglycemia (in the past 12 months), and, for respondents reporting recent and/or severe hypoglycemia, lifetime experience of vehicle crashes or severe medical injury, daily debilitating worry, and withdrawal from driving, exercise, sex, and going outside of the home to avoid hypoglycemia and consequences. Secondary outcomes included measures of research engagement. RESULTS: Of 613 respondents (24.3% of app users), 49.1% reported more than 4 episodes of "going low" in the past 2 weeks and 29.2% reported 1 or more severe low in the past year; 16.6% reported both more than 4 recent low episodes and 1 or more severe event in the past year. Harms were common, including daily debilitating worry (45.8%), vehicle crash or injury (15.0%), and withdrawal from exercise, driving, leaving home, and having sex (54.0%, 37.4%, 24.8%, and 22.7%, respectively). Of all respondents, 54.2% reported multiple harms, the risks for which were highest (73.7%) among respondents with a past-year severe event (odds ratio, 2.39; 95% CI, 1.60-3.58; P < .001 controlling for frequent recent low episodes and demographic and disease factors). Engagement was high, with 96.6% of the sample permitting recontact for research and 31.7% posting personal study data on their app profile page; 40.5% of 2825 unique page views of research-related materials published on the community site involved views of returned research results. CONCLUSIONS AND RELEVANCE: Participatory surveillance of hypoglycemia in an online diabetes social network enables characterization of patient-centered harms in a community sample and bidirectional communication with affected persons, augmenting traditional surveillance.
Murthy S, Mandl KD, Bourgeois F. Analysis of pediatric clinical drug trials for neuropsychiatric conditions. Pediatrics. 2013;131:1125-31.Abstract
BACKGROUND AND OBJECTIVE: Neuropsychiatric conditions represent a large and increasing disease burden in children. A number of drugs are available for the treatment of these conditions, but most drugs have not been adequately tested in children, and off-label drug use remains widespread. We sought to define and quantify recent and ongoing clinical research on the use of neuropsychiatric drugs in children. METHODS: Drug trials registered in ClinicalTrials.gov between 2006 and 2011 and studying neuropsychiatric conditions were selected and classified based on the drug's Food and Drug Administration (FDA) approval status in children. We measured the proportion of trials seeking to expand the use of a drug to pediatric patients and the proportion of available drugs studied in children. RESULTS: Only 10% of neuropsychiatric trials focused on children. Of 303 drugs studied in both pediatric and adult populations, 90% lacked FDA approval in children and 97% were not approved in children for the indication studied. However, only 19% of all neuropsychiatric drugs were under study in pediatric populations, with as few as 8% of either antidepressant or antipsychotic drugs. Overall, 76% of pediatric drug trials examined a drug previously unapproved in children and 26% explored the use of a drug for a new indication. CONCLUSIONS: Despite the rising prevalence of neuropsychiatric disease and the paucity of FDA-approved pediatric drugs, only a small proportion of trials focus on pediatric populations and these trials cover only a fraction of available drugs. This deficiency is most pronounced for depression and schizophrenia.
Natter MD, Quan J, Ortiz DM, Bousvaros A, Ilowite NT, Inman CJ, Marsolo K, McMurry AJ, Sandborg CI, Schanberg LE, et al. An i2b2-based, generalizable, open source, self-scaling chronic disease registry. J Am Med Inform Assoc. 2013;20:172-9.Abstract
OBJECTIVE: Registries are a well-established mechanism for obtaining high quality, disease-specific data, but are often highly project-specific in their design, implementation, and policies for data use. In contrast to the conventional model of centralized data contribution, warehousing, and control, we design a self-scaling registry technology for collaborative data sharing, based upon the widely adopted Integrating Biology & the Bedside (i2b2) data warehousing framework and the Shared Health Research Information Network (SHRINE) peer-to-peer networking software. MATERIALS AND METHODS: Focusing our design around creation of a scalable solution for collaboration within multi-site disease registries, we leverage the i2b2 and SHRINE open source software to create a modular, ontology-based, federated infrastructure that provides research investigators full ownership and access to their contributed data while supporting permissioned yet robust data sharing. We accomplish these objectives via web services supporting peer-group overlays, group-aware data aggregation, and administrative functions. RESULTS: The 56-site Childhood Arthritis & Rheumatology Research Alliance (CARRA) Registry and 3-site Harvard Inflammatory Bowel Diseases Longitudinal Data Repository now utilize i2b2 self-scaling registry technology (i2b2-SSR). This platform, extensible to federation of multiple projects within and between research networks, encompasses >6000 subjects at sites throughout the USA. DISCUSSION: We utilize the i2b2-SSR platform to minimize technical barriers to collaboration while enabling fine-grained control over data sharing. CONCLUSIONS: The implementation of i2b2-SSR for the multi-site, multi-stakeholder CARRA Registry has established a digital infrastructure for community-driven research data sharing in pediatric rheumatology in the USA. We envision i2b2-SSR as a scalable, reusable solution facilitating interdisciplinary research across diseases.
Cassa CA, Miller RA, Mandl KD. A novel, privacy-preserving cryptographic approach for sharing sequencing data. J Am Med Inform Assoc. 2013;20:69-76.Abstract
OBJECTIVE: DNA samples are often processed and sequenced in facilities external to the point of collection. These samples are routinely labeled with patient identifiers or pseudonyms, allowing for potential linkage to identity and private clinical information if intercepted during transmission. We present a cryptographic scheme to securely transmit externally generated sequence data which does not require any patient identifiers, public key infrastructure, or the transmission of passwords. MATERIALS AND METHODS: This novel encryption scheme cryptographically protects participant sequence data using a shared secret key that is derived from a unique subset of an individual's genetic sequence. This scheme requires access to a subset of an individual's genetic sequence to acquire full access to the transmitted sequence data, which helps to prevent sample mismatch. RESULTS: We validate that the proposed encryption scheme is robust to sequencing errors, population uniqueness, and sibling disambiguation, and provides sufficient cryptographic key space. DISCUSSION: Access to a set of an individual's genotypes and a mutually agreed cryptographic seed is needed to unlock the full sequence, which provides additional sample authentication and authorization security. We present modest fixed and marginal costs to implement this transmission architecture. CONCLUSIONS: It is possible for genomics researchers who sequence participant samples externally to protect the transmission of sequence data using unique features of an individual's genetic sequence.
Berry JG, Hall M, Hall DE, Kuo DZ, Cohen E, Agrawal R, Mandl KD, Clifton H, Neff J. Inpatient growth and resource use in 28 children's hospitals: a longitudinal, multi-institutional study. JAMA Pediatr. 2013;167:170-7.Abstract
OBJECTIVE: To compare inpatient resource use trends for healthy children and children with chronic health conditions of varying degrees of medical complexity. DESIGN: Retrospective cohort analysis. SETTING: Twenty-eight US children's hospitals. PATIENTS: A total of 1 526 051 unique patients hospitalized from January 1, 2004, through December 31, 2009, who were assigned to 1 of 5 chronic condition groups using 3M's Clinical Risk Group software. INTERVENTION: None. MAIN OUTCOME MEASURES: Trends in the number of patients, hospitalizations, hospital days, and charges analyzed with linear regression. RESULTS: Between 2004 and 2009, hospitals experienced a greater increase in the number of children hospitalized with vs without a chronic condition (19.2% vs 13.7% cumulative increase, P < .001). The greatest cumulative increase (32.5%) was attributable to children with a significant chronic condition affecting 2 or more body systems, who accounted for 19.2% (n = 63 203) of patients, 27.2% (n = 111 685) of hospital discharges, 48.9% (n = 1.1 million) of hospital days, and 53.2% ($9.2 billion) of hospital charges in 2009. These children had a higher percentage of Medicaid use (56.5% vs 49.7%; P < .001) compared with children without a chronic condition. Cerebral palsy (9179 [14.6%]) and asthma (13 708 [21.8%]) were the most common primary diagnosis and comorbidity, respectively, observed among these patients. CONCLUSIONS: Patients with a chronic condition increasingly used more resources in a group of children's hospitals than patients without a chronic condition. The greatest growth was observed in hospitalized children with chronic conditions affecting 2 or more body systems. Children's hospitals must ensure that their inpatient care systems and payment structures are equipped to meet the protean needs of this important population of children.
Cassa CA, Chunara R, Mandl K, Brownstein JS. Twitter as a sentinel in emergency situations: lessons from the Boston marathon explosions. PLoS Curr. 2013;5.Abstract
Immediately following the Boston Marathon attacks, individuals near the scene posted a deluge of data to social media sites. Previous work has shown that these data can be leveraged to provide rapid insight during natural disasters, disease outbreaks and ongoing conflicts that can assist in the public health and medical response. Here, we examine and discuss the social media messages posted immediately after and around the Boston Marathon bombings, and find that specific keywords appear frequently prior to official public safety and news media reports. Individuals immediately adjacent to the explosions posted messages within minutes via Twitter which identify the location and specifics of events, demonstrating a role for social media in the early recognition and characterization of emergency events. *Christopher Cassa and Rumi Chunara contributed equally to this work.
Mandl KD, McNabb M, Marks N, Weitzman ER, Kelemen S, Eggleston EM, Quinn M. Participatory surveillance of diabetes device safety: a social media-based complement to traditional FDA reporting. J Am Med Inform Assoc. 2013.Abstract
BACKGROUND AND OBJECTIVE: Malfunctions or poor usability of devices measuring glucose or delivering insulin are reportable to the FDA. Manufacturers submit 99.9% of these reports. We test online social networks as a complementary source to traditional FDA reporting of device-related adverse events. METHODS: Participatory surveillance of members of a non-profit online social network, TuDiabetes.org, from October 2011 to September 2012. Subjects were volunteers from a group within TuDiabetes, actively engaged online in participatory surveillance. They used the free TuAnalyze app, a privacy-preserving method to report detailed clinical information, available through the network. Network members were polled about finger-stick blood glucose monitors, continuous glucose monitors, and insulin delivery devices, including insulin pumps and insulin pens. RESULTS: Of 549 participants, 75 reported device-related adverse events, nearly half (48.0%) requiring intervention from another person to manage the event. Only three (4.0%) of these were reported by participants to the FDA. All TuAnalyze reports contained outcome information compared with 22% of reports to the FDA. Hypoglycemia and hyperglycemia were experienced by 48.0% and 49.3% of participants, respectively. DISCUSSION: Members of an online community readily engaged in participatory surveillance. While polling distributed online populations does not yield generalizable, denominator-based rates, this approach can characterize risk within online communities using a bidirectional communication channel that enables reach-back and intervention. CONCLUSIONS: Engagement of distributed communities in social networks is a viable complementary approach to traditional public health surveillance for adverse events related to medical devices.
Zimolzak AJ, Spettell CM, Fernandes J, Fusaro VA, Palmer NP, Saria S, Kohane IS, Jonikas MA, Mandl KD. Early detection of poor adherers to statins: applying individualized surveillance to pay for performance. PLoS One. 2013;8:e79611.Abstract
BACKGROUND: Medication nonadherence costs $300 billion annually in the US. Medicare Advantage plans have a financial incentive to increase medication adherence among members because the Centers for Medicare and Medicaid Services (CMS) now awards substantive bonus payments to such plans, based in part on population adherence to chronic medications. We sought to build an individualized surveillance model that detects early which beneficiaries will fall below the CMS adherence threshold. METHODS: This was a retrospective study of over 210,000 beneficiaries initiating statins, in a database of private insurance claims, from 2008-2011. A logistic regression model was constructed to use statin adherence from initiation to day 90 to predict beneficiaries who would not meet the CMS measure of proportion of days covered 0.8 or above, from day 91 to 365. The model controlled for 15 additional characteristics. In a sensitivity analysis, we varied the number of days of adherence data used for prediction. RESULTS: Lower adherence in the first 90 days was the strongest predictor of one-year nonadherence, with an odds ratio of 25.0 (95% confidence interval 23.7-26.5) for poor adherence at one year. The model had an area under the receiver operating characteristic curve of 0.80. Sensitivity analysis revealed that predictions of comparable accuracy could be made only 40 days after statin initiation. When members with 30-day supplies for their first statin fill had predictions made at 40 days, and members with 90-day supplies for their first fill had predictions made at 100 days, poor adherence could be predicted with 86% positive predictive value. CONCLUSIONS: To preserve their Medicare Star ratings, plan managers should identify or develop effective programs to improve adherence. An individualized surveillance approach can be used to target members who would most benefit, recognizing the tradeoff between improved model performance over time and the advantage of earlier detection.
Dunn AG, Mandl KD, Coiera E, Bourgeois FT. The effects of industry sponsorship on comparator selection in trial registrations for neuropsychiatric conditions in children. PLoS One. 2013;8:e84951.Abstract
Pediatric populations continue to be understudied in clinical drug trials despite the increasing use of pharmacotherapy in children, particularly with psychotropic drugs. Most pertinent to the clinical selection of drug interventions are trials directly comparing drugs against other drugs. The aim was to measure the prevalence of active drug comparators in neuropsychiatric drug trials in children and identify the effects of funding source on comparator selection. We analyzed the selection of drugs and drug comparisons in clinical trials registered between January 2006 and May 2012. Completed and ongoing interventional trials examining treatments for six neuropsychiatric conditions in children were included. Networks of drug comparisons for each condition were constructed using information about the trial study arms. Of 421 eligible trial registrations, 228 (63,699 participants) were drug trials addressing ADHD (106 trials), autism spectrum disorders (47), unipolar depression (16), seizure disorders (38), migraines and other headaches (15), or schizophrenia (11). Active drug comparators were used in only 11.0% of drug trials while 44.7% used a placebo control and 44.3% no drug or placebo comparator. Even among conditions with well-established pharmacotherapeutic options, almost all drug interventions were compared to a placebo. Active comparisons were more common among trials without industry funding (17% vs. 8%, p=0.04). Trials with industry funding differed from non-industry trials in terms of the drugs studied and the comparators selected. For 73% (61/84) of drugs and 90% (19/21) of unique comparisons, trials were funded exclusively by either industry or non-industry. We found that industry and non-industry differed when choosing comparators and active drug comparators were rare for both groups. This gap in pediatric research activity limits the evidence available to clinicians treating children and suggests a need to reassess the design and funding of pediatric trials in order to optimize the information derived from pediatric participation in clinical trials.
Bosl W, Mandel J, Jonikas M, Ramoni RB, Kohane IS, Mandl KD. Scalable decision support at the point of care: a substitutable electronic health record app for monitoring medication adherence. Interact J Med Res. 2013;2:e13.Abstract
BACKGROUND: Non-adherence to prescribed medications is a serious health problem in the United States, costing an estimated $100 billion per year. While poor adherence should be addressable with point of care health information technology, integrating new solutions with existing electronic health records (EHR) systems require customization within each organization, which is difficult because of the monolithic software design of most EHR products. OBJECTIVE: The objective of this study was to create a published algorithm for predicting medication adherence problems easily accessible at the point of care through a Web application that runs on the Substitutable Medical Apps, Reusuable Technologies (SMART) platform. The SMART platform is an emerging framework that enables EHR systems to behave as "iPhone like platforms" by exhibiting an application programming interface for easy addition and deletion of third party apps. The app is presented as a point of care solution to monitoring medication adherence as well as a sufficiently general, modular application that may serve as an example and template for other SMART apps. METHODS: The widely used, open source Django framework was used together with the SMART platform to create the interoperable components of this app. Django uses Python as its core programming language. This allows statistical and mathematical modules to be created from a large array of Python numerical libraries and assembled together with the core app to create flexible and sophisticated EHR functionality. Algorithms that predict individual adherence are derived from a retrospective study of dispensed medication claims from a large private insurance plan. Patients' prescription fill information is accessed through the SMART framework and the embedded algorithms compute adherence information, including predicted adherence one year after the first prescription fill. Open source graphing software is used to display patient medication information and the results of statistical prediction of future adherence on a clinician-facing Web interface. RESULTS: The user interface allows the physician to quickly review all medications in a patient record for potential non-adherence problems. A gap-check and current medication possession ratio (MPR) threshold test are applied to all medications in the record to test for current non-adherence. Predictions of 1-year non-adherence are made for certain drug classes for which external data was available. Information is presented graphically to indicate present non-adherence, or predicted non-adherence at one year, based on early prescription fulfillment patterns. The MPR Monitor app is installed in the SMART reference container as the "MPR Monitor", where it is publically available for use and testing. MPR is an acronym for Medication Possession Ratio, a commonly used measure of adherence to a prescribed medication regime. This app may be used as an example for creating additional functionality by replacing statistical and display algorithms with new code in a cycle of rapid prototyping and implementation or as a framework for a new SMART app. CONCLUSIONS: The MPR Monitor app is a useful pilot project for monitoring medication adherence. It also provides an example that integrates several open source software components, including the Python-based Django Web framework and python-based graphics, to build a SMART app that allows complex decision support methods to be encapsulated to enhance EHR functionality.
2012
Dunn AG, Day RO, Mandl KD, Coiera E. Learning from hackers: open-source clinical trials. Sci Transl Med. 2012;4:132cm5.Abstract
Open sharing of clinical trial data has been proposed as a way to address the gap between the production of clinical evidence and the decision-making of physicians. A similar gap was addressed in the software industry by their open-source software movement. Here, we examine how the social and technical principles of the movement can guide the growth of an open-source clinical trial community.
Reis BY, Olson KL, Tian L, Bohn RL, Brownstein JS, Park PJ, Cziraky MJ, Wilson MD, Mandl KD. A pharmacoepidemiological network model for drug safety surveillance: statins and rhabdomyolysis. Drug Saf. 2012;35:395-406.Abstract
BACKGROUND: Recent withdrawals of major drugs have highlighted the critical importance of drug safety surveillance in the postmarketing phase. Limitations of spontaneous report data have led drug safety professionals to pursue alternative postmarketing surveillance approaches based on healthcare administrative claims data. These data are typically analysed by comparing the adverse event rates associated with a drug of interest to those of a single comparable reference drug. OBJECTIVE: The aim of this study was to determine whether adverse event detection can be improved by incorporating information from multiple reference drugs. We developed a pharmacological network model that implemented this approach and evaluated its performance. METHODS: We studied whether adverse event detection can be improved by incorporating information from multiple reference drugs, and describe two approaches for doing so. The first, reported previously, combines a set of related drugs into a single reference cohort. The second is a novel pharmacoepidemiological network model, which integrates multiple pair-wise comparisons across an entire set of related drugs into a unified consensus safety score for each drug. We also implemented a single reference drug approach for comparison with both multi-drug approaches. All approaches were applied within a sequential analysis framework, incorporating new information as it became available and addressing the issue of multiple testing over time. We evaluated all these approaches using statin (HMG-CoA reductase inhibitors) safety data from a large healthcare insurer in the US covering April 2000 through March 2005. RESULTS: We found that both multiple reference drug approaches offer earlier detection (6-13 months) than the single reference drug approach, without triggering additional false positives. CONCLUSIONS: Such combined approaches have the potential to be used with existing healthcare databases to improve the surveillance of therapeutics in the postmarketing phase over single-comparator methods. The proposed network approach also provides an integrated visualization framework enabling decision makers to understand the key high-level safety relationships amongst a group of related drugs.
Cassa CA, Savage SK, Taylor PL, Green RC, McGuire AL, Mandl KD. Disclosing pathogenic genetic variants to research participants: quantifying an emerging ethical responsibility. Genome Res. 2012;22:421-8.Abstract
There is an emerging consensus that when investigators obtain genomic data from research participants, they may incur an ethical responsibility to inform at-risk individuals about clinically significant variants discovered during the course of their research. With whole-exome sequencing becoming commonplace and the falling costs of full-genome sequencing, there will be an increasingly large number of variants identified in research participants that may be of sufficient clinical relevance to share. An explicit approach to triaging and communicating these results has yet to be developed, and even the magnitude of the task is uncertain. To develop an estimate of the number of variants that might qualify for disclosure, we apply recently published recommendations for the return of results to a defined and representative set of variants and then extrapolate these estimates to genome scale. We find that the total number of variants meeting the threshold for recommended disclosure ranges from 3955-12,579 (3.79%-12.06%, 95% CI) in the most conservative estimate to 6998-17,189 (6.69%-16.48%, 95% CI) in an estimate including variants with variable disease expressivity. Additionally, if the growth rate from the previous 4 yr continues, we estimate that the total number of disease-associated variants will grow 37% over the next 4 yr.