Devan Kansagara

Risk Prediction Models for Hospital Readmission: A Systematic Review

CONTEXT Predicting hospital readmission risk is of great interest to identify which patients would benefit most from care transition interventions, as well as to risk-adjust readmission rates for the purposes of hospital comparison. OBJECTIVE To summarize validated readmission risk prediction models, describe their performance, and assess suitability for clinical or administrative use. DATA SOURCES AND STUDY SELECTION The databases of MEDLINE, CINAHL, and the Cochrane Library were searched from inception through March 2011, the EMBASE database was searched through August 2011, and hand searches were performed of the retrieved reference lists. Dual review was conducted to identify studies published in the English language of prediction models tested with medical patients in both derivation and validation cohorts. DATA EXTRACTION Data were extracted on the population, setting, sample size, follow-up interval, readmission rate, model discrimination and calibration, type of data used, and timing of data collection. DATA SYNTHESIS Of 7843 citations reviewed, 30 studies of 26 unique models met the inclusion criteria. The most common outcome used was 30-day readmission; only 1 model specifically addressed preventable readmissions. Fourteen models that relied on retrospective administrative data could be potentially used to risk-adjust readmission rates for hospital comparison; of these, 9 were tested in large US populations and had poor discriminative ability (c statistic range: 0.55-0.65). Seven models could potentially be used to identify high-risk patients for intervention early during a hospitalization (c statistic range: 0.56-0.72), and 5 could be used at hospital discharge (c statistic range: 0.68-0.83). Six studies compared different models in the same population and 2 of these found that functional and social variables improved model discrimination. Although most models incorporated variables for medical comorbidity and use of prior medical services, few examined variables associated with overall health and function, illness severity, or social determinants of health. CONCLUSIONS Most current readmission risk prediction models that were designed for either comparative or clinical purposes perform poorly. Although in certain settings such models may prove useful, efforts to improve their performance are needed as use becomes more widespread.

Management of Chronic Insomnia Disorder in Adults: A Clinical Practice Guideline From the American College of Physicians

DESCRIPTION The American College of Physicians (ACP) developed this guideline to present the evidence and provide clinical recommendations on the management of chronic insomnia disorder in adults. METHODS This guideline is based on a systematic review of randomized, controlled trials published in English from 2004 through September 2015. Evaluated outcomes included global outcomes assessed by questionnaires, patient-reported sleep outcomes, and harms. The target audience for this guideline includes all clinicians, and the target patient population includes adults with chronic insomnia disorder. This guideline grades the evidence and recommendations by using the ACP grading system, which is based on the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach. RECOMMENDATION 1 ACP recommends that all adult patients receive cognitive behavioral therapy for insomnia (CBT-I) as the initial treatment for chronic insomnia disorder. (Grade: strong recommendation, moderate-quality evidence). RECOMMENDATION 2 ACP recommends that clinicians use a shared decision-making approach, including a discussion of the benefits, harms, and costs of short-term use of medications, to decide whether to add pharmacological therapy in adults with chronic insomnia disorder in whom cognitive behavioral therapy for insomnia (CBT-I) alone was unsuccessful. (Grade: weak recommendation, low-quality evidence).

Intensive Insulin Therapy in Hospitalized Patients: A Systematic Review

BACKGROUND The benefits and harms of intensive insulin therapy (IIT) titrated to strict glycemic targets in hospitalized patients remain uncertain. PURPOSE To evaluate the benefits and harms of IIT in hospitalized patients. DATA SOURCES MEDLINE and Cochrane Database of Systematic Reviews from 1950 to January 2010, reference lists, experts, and unpublished sources. STUDY SELECTION English-language randomized, controlled trials comparing protocols titrated to strict or less strict glycemic targets. DATA EXTRACTION Two reviewers independently abstracted data from each study on sample, setting, glycemic control interventions, glycemic targets, mean glucose levels achieved, and outcomes. Results were grouped by patient population or setting. A random-effects model was used to combine trial data on short-term mortality (≤28 days), long-term mortality (90 or 180 days), infection, length of stay, and hypoglycemia. The Grading of Recommendations Assessment, Development, and Evaluation system was used to rate the overall body of evidence for each outcome. DATA SYNTHESIS In a meta-analysis of 21 trials in intensive care unit, perioperative care, myocardial infarction, and stroke or brain injury settings, IIT did not affect short-term mortality (relative risk, 1.00 [95% CI, 0.94 to 1.07]). No consistent evidence showed that IIT reduced long-term mortality, infection rates, length of stay, or the need for renal replacement therapy. No evidence of benefit from IIT was reported in any hospital setting, although the best evidence for lack of benefit was in intensive care unit settings. Data combined from 10 trials showed that IIT was associated with a high risk for severe hypoglycemia (relative risk, 6.00 [CI, 4.06 to 8.87]; P < 0.001). Risk for IIT-associated hypoglycemia was increased in all hospital settings. LIMITATIONS Methodological shortcomings and inconsistencies limit the data in perioperative care, myocardial infarction, and stroke or brain injury settings. Differences in insulin protocols and patient and hospital characteristics may affect generalizability across treatment settings. CONCLUSION No consistent evidence demonstrates that IIT targeted to strict glycemic control compared with less strict glycemic control improves health outcomes in hospitalized patients. Furthermore, IIT is associated with an increased risk for severe hypoglycemia. PRIMARY FUNDING SOURCE U.S. Department of Veterans Affairs Health Services Research and Development Service.

Screening adults for type 2 diabetes: a review of the evidence for the U.S. Preventive Services Task Force.

The 2002 National Health and Nutrition Examination Survey estimated that 19.3 million U.S. adults (9.3% of the total U.S. population) had diabetes mellitus, one third of whom had undiagnosed diabetes (1). In addition, 26.0% had impaired fasting glucose and impaired glucose tolerance was even more prevalent (2). The prevalence of diagnosed diabetes is increasing, particularly among obese individuals (1, 3). The risk for death among persons with diabetes is about twice that of persons without diabetes, and cardiovascular events account for more than three fourths of these deaths (4). Type 2 diabetes often goes undiagnosed for many years because hyperglycemia develops gradually and may not produce symptoms (3, 5). Persons with diabetes are at increased risk for microvascular and macrovascular complications, and duration of diabetes and degree of hyperglycemia are associated with an increased risk for microvascular complications (69). The prevalence of macrovascular complications is elevated in persons with prediabetes (defined as impaired fasting glucose, impaired glucose tolerance, or both) and in persons with newly diagnosed diabetes (1018). A substantial proportion of persons presenting with a new cardiovascular event have undiagnosed diabetes or prediabetes (10, 1923). Several recent observational studies and a meta-analysis suggest an association between chronic hyperglycemia and cardiovascular disease and stroke (2427). Diabetes has a long preclinical phase, estimated at 10 to 12 years on the basis of the progression of microvascular complications (28), and valid and reliable tests can detect type 2 diabetes during this asymptomatic period (29). The previous evidence review for the U.S. Preventive Services Task Force (USPSTF) (29, 30) suggested that screening is justified if the ensuing treatments offer incremental net benefits compared with treatment at the time of clinical presentation. In 2003, the USPSTF concluded that the evidence was insufficient to recommend for or against routinely screening asymptomatic adults for type 2 diabetes or prediabetes, but it did recommend screening for adults with hypertension or hyperlipidemia (31). This review summarizes evidence that has become available since the previous report to inform an update of the 2003 USPSTF recommendations on screening for type 2 diabetes and prediabetes. Methods The methods of the USPSTF evidence reviews are fully detailed elsewhere (32). The analytic framework (Appendix Figure 1) focuses on decreasing the risk for complications from type 2 diabetes as a result of screening for diabetes. We did not consider secondary prevention studies that exclusively enrolled persons with known cardiovascular disease, because we considered those persons to have a potential preexisting diabetes complication. Appendix Figure 1. Analytic framework. KQ= key question. We searched MEDLINE and the Cochrane Library for relevant English-language systematic reviews and studies published between March 2001 (6 months before the end date of the previous search) and July 2007 related to 5 key questions. We also examined the reference lists of included studies and ClinicalTrials.gov for relevant trials. We evaluated all studies included in relevant systematic reviews for potential inclusion. We included randomized, controlled trials (RCTs) and observational studies that examined the effectiveness or adverse effects of screening and diagnosis of type 2 diabetes. We used RCTs to assess the effectiveness of diabetes and prediabetes treatments. For diabetes interventions, we included trials with patients who had disease for 1 year or less, as well as trials comparing outcomes among diabetic and nondiabetic populations. We used good-quality systematic reviews to assess the adverse effects of treatment. Search strategies are available in the full evidence report, which can be found at www.ahrq.gov/clinic/uspstf/uspstopics.htm. An investigator screened titles and abstracts, and a random sample of 1500 titles and abstracts was dual reviewed. Two reviewers examined the full text of potentially relevant articles to achieve consensus on inclusion (Figure). Data were abstracted by one investigator and checked by another. We assessed internal validity of individual trials by examining factors that might introduce bias: adequate randomization, allocation concealment, baseline comparability of participants, blinding, and loss to follow-up. We rated studies as good, fair, or poor quality by using standard USPSTF criteria (32). We rated systematic reviews on the basis of established criteria, and we included only good-quality reviews (33, 34). We assessed potential applicability of individual studies to primary care practice on the basis of the methods of participant recruitment and selection. Figure. Study flow diagram. KQ= key question. *Cochrane databases were the Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and Database of Abstracts of Reviews of Effects. Other sources were reference lists and expert referrals. We identified studies that modeled screening interventions from our main search, as well as from a recent, good-quality systematic review of screening for type 2 diabetes by the National Health Service Research and Development Health Technology Assessment Programme (35). We independently abstracted the relevant studies included in that report and relied on their extensive assessments of model quality. We performed a qualitative synthesis of abstracted data that were generally too heterogeneous for quantitative pooling, except for estimates of the effect of pharmacotherapeutic or lifestyle interventions on diabetes incidence in prediabetic populations. We calculated these pooled estimates by using a hazard ratio and its SE from Cox regression; either a rate ratio or a risk ratio was calculated when a hazard ratio was not reported (3638). We tested for statistical heterogeneity with the standard chi-square test and obtained the overall estimates of relative risk by using a random-effects model (39). Role of the Funding Source This study was funded by the Agency for Healthcare Research and Quality (AHRQ) under a contract to support the work of the USPSTF. Agency staff and USPSTF members participated in the initial scope of this work and reviewed interim analyses and the final report. A draft version was distributed to content experts for review. Agency approval was required before this manuscript could be submitted for publication, but the authors are solely responsible for the content and the decision to submit it for publication. Results Key Question 1 Is there direct evidence that systematic screening for type 2 diabetes, impaired fasting glucose, or impaired glucose tolerance among asymptomatic adults improves health outcomes? We identified no RCTs examining the effectiveness of a screening program for type 2 diabetes. A small, good-quality, casecontrol study did not find benefit from screening when microvascular complications were considered (40). Limited data from 2 cross-sectional studies did not provide good-quality, direct evidence of the effectiveness of screening for type 2 diabetes in either targeted or general populations (41, 42). Of modeling studies identified (35, 4348), 2 recent high-quality studies suggested that targeted screening for type 2 diabetes among persons with hypertension may be relatively cost-effective when macrovascular benefits of optimal blood pressure control are considered (35, 47), older persons benefitted more than younger persons (35, 47), and screening obese persons was more cost-effective than mass screening (35). The ADDITION (Anglo-Danish-Dutch Study of Intensive Treatment in People with Screen Detected Diabetes in Primary Care) study (49), currently in progress, may shed light on differences in baseline characteristics and long-term health outcomes between persons with screening-detected diabetes and those who present with symptoms. Key Question 2 Does beginning treatment of type 2 diabetes early as a result of screening provide an incremental benefit in health outcomes compared with initiating treatment after clinical diagnosis? We identified no studies that directly explored this question by comparing treatment effects between persons with screening-detected versus clinically detected diabetes, nor did we identify new studies reporting treatment effects in an exclusively screening-detected or recently diagnosed diabetes cohort. Because of the lack of direct evidence, we examined intervention studies comparing treatment effects in diabetic versus nondiabetic populations (5061) to address the question: Would early knowledge of a diabetes diagnosis prompt a change in clinical management? Tight Glycemic Control No new, completed studies have examined the effect of glycemic control strategies in persons with newly diagnosed type 2 diabetes since the previous USPSTF review (29). The UKPDS (United Kingdom Prospective Diabetes Study) (62) remains the largest and most influential trial of intensive glycemic control in persons with newly diagnosed, mainly clinically detected, type 2 diabetes. In the UKPDS, persons assigned to intensive glycemic control had a 25% reduction (95% CI, 7% to 40%) in microvascular complications, mostly due to a reduced need for retinal photocoagulation, as well as a nonsignificant 16% relative risk reduction (CI, 71% to 100%) of myocardial infarction (62). The UKPDS investigators estimated that 19.6 persons (CI, 10 to 500 persons) would need to be intensively treated for 10 years to prevent 1 person from developing any single clinical end point (62). A recent meta-analysis combined results from the UKPDS and other older trials examined in the last USPSTF review, and it concluded that tight glycemic control resulted in a modest reduction of macrovascular events, particularly peripheral vascular and cerebrovascular events, in persons with type 2 diabetes (combined incidenc

Screening for hepatocellular carcinoma in chronic liver disease: a systematic review.

Hepatocellular carcinoma (HCC) incidence and mortality have increased internationally over the past 4 decades (1, 2), with localized tumors accounting for most of the increase (3). The rationale for screening is that imaging tests, such as ultrasonography, may identify patients with early-stage HCC (4), and several potential options exist for treating patients with early-stage HCC, including liver transplantation, radiofrequency ablation, and liver resection (5). Several professional societies currently recommend HCC screening using imaging studies and tumor markers, primarily in patients at higher risk for HCC due to chronic hepatitis B or cirrhosis (57). However, recommendations for HCC screening remain controversial, in part because of concerns over the quality and paucity of existing evidence and because concerns about overdiagnosis and patient harms have been raised in other cancer screening programs (812). We conducted a systematic review of the published literature to better understand the incremental benefits and harms of routine HCC screening compared with clinical diagnosis. Methods This manuscript is part of a larger report commissioned by the Veterans Health Administration (13). A protocol describing the review plan was posted to a public Web site before the study was initiated (14). The analytic framework that guided this review was developed in collaboration with a panel of technical experts and is provided in Figure 1 of Supplement 1. Supplement 1. Figures Data Sources and Searches We searched MEDLINE, PsycINFO, the Cochrane Central Register of Controlled Trials, the Cochrane Database of Systematic Reviews, and ClinicalTrials.gov from database inception to June 2013. We updated the MEDLINE, PsycINFO, and ClinicalTrials.gov searches in April 2014. The detailed search strategy is provided in Supplement 2. We obtained additional articles from systematic reviews, reference lists of pertinent studies, reviews, and editorials and by consulting technical advisors. Supplement 2. Search Strategy Study Selection Detailed inclusion and exclusion criteria are provided in Supplement 3. We included English-language, controlled clinical trials and observational studies that assessed the effects of screening on HCC-specific and all-cause mortality in adult populations. We used the term screening to include any surveillance or screening program in which specific tests (ultrasonography, computed tomography, magnetic resonance imaging, or -fetoprotein measurement) were performed explicitly to detect HCC in asymptomatic patients. Studies had to include a comparison group of patients who did not have routine screening. We excluded observational studies that did not consider important confounding factors, such as age, sex, and liver disease severity. Because we anticipated few clinical trials comparing screening versus no screening, we also included trials comparing frequencies of screening. We included studies of any population with chronic liver disease with or without cirrhosis but excluded studies of patients with prior HCC. We also searched for systematic reviews and primary studies that focused on potential harms of HCC screening. Supplement 3. Inclusion/Exclusion Criteria Seven investigators reviewed the titles and abstracts of citations identified from literature searches. If at least 1 reviewer indicated that a citation may be relevant, a second reviewer screened the citation for concordance. Two reviewers independently assessed the full-text articles for inclusion using the eligibility criteria in Supplement 3. Disagreements were resolved through consensus. Data Extraction and Quality Assessment From each study, we abstracted study design, objectives, setting, population characteristics (including sex, age, race or ethnicity, and liver disease cause and severity), patient eligibility and exclusion criteria, number of patients, years of enrollment, method and frequency of screening, adjusted and unadjusted mortality, and adverse events. A second author checked each entry for accuracy. Two reviewers independently assessed the quality of each trial by using a tool developed by the Cochrane Collaboration (15). We resolved disagreements through discussion. Each trial was given an overall summary assessment of low, high, or unclear risk of bias. Two reviewers graded the strength of evidence for outcomes by using published criteria that consider the consistency, coherence, directness, and applicability of a body of evidence as well as the internal validity of individual studies (16). We adapted existing tools to assess the quality of observational studies (1719). We do not report an overall summary assessment for observational studies because there are no validated criteria for doing so. Data Synthesis and Analysis We qualitatively synthesized the evidence on the benefits and harms of HCC screening. Clinical heterogeneity and the small number of trials precluded a meta-analysis of the findings. Role of the Funding Source The U.S. Department of Veterans Affairs Quality Enhancement Research Initiative supported this review but had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. Results The electronic and manual searches yielded 13801 total citations, from which we identified 286 potentially relevant full-text articles. Twenty-two primary studies contained primary data relevant to the efficacy of HCC screening and met our inclusion criteria (Figure). Figure. Summary of evidence search and selection. RCT = randomized, controlled trial. Effects of Screening on Mortality Two trials and 18 observational studies provided very-low-strength evidence from which to draw conclusions about the mortality effects of HCC screening compared with no screening. The trials had substantial methodological flaws that threatened their internal validity, and their findings have limited applicability beyond the patient population with hepatitis B. The observational studies, most of which included patients with cirrhosis and hepatitis B, hepatitis C, or alcoholic liver disease, showed that screening detects patients with earlier-stage disease, who more frequently receive curative therapy. However, it is impossible to say whether the longer survival in patients with screen-detected disease was a true effect of screening or reflects lead- and length-time biases inherent to all observational studies, as well as selection biases that were common in many of the studies. Randomized, Controlled Trials Two community-based trials compared the effects on mortality of screening versus no screening (20, 21). Both were conducted in China in areas with high prevalence of HCC, and most participants had hepatitis B with or without cirrhosis (Table 1 of Supplement 4). One cluster randomized trial recruited screening group participants (n= 9757) from 1993 to 1995 and offered them serum -fetoprotein testing and ultrasonography every 6 months. Participants in the control group (n= 9443) were not made aware of the study nor actively followed. Death from HCC occurred less frequently in the screening group (83.2 vs. 131.5 per 100000 person-years; rate ratio, 0.63 [95% CI, 0.41 to 0.98]). Supplement 4. Tables However, the trial had several serious methodological limitations that gave it a high risk of bias (Table 2 of Supplement 4). One major concern is whether patients in both groups had the same risk for HCC. There is no information about randomization technique or allocation concealment and very little information about the baseline characteristics of the 2 groups, which is especially important in cluster randomized trials. Another concern is that weak methods used to ascertain the outcome measuredeath from HCCcould have introduced bias. If deaths were underreported in the control group, results could have been biased toward the null. On the other hand, if outcome adjudicators were not blinded, more deaths in the control group could have been misclassified as HCC-related, especially because the symptoms that define stage III HCC (cachexia, jaundice, and ascites) overlap substantially with symptoms of end-stage liver disease and no data were provided about liver disease severity in either group. Selective reporting and analysis of favorable outcomes were other concerns. Although the authors reported that vital status was available for all patients, overall mortality was not reported and there was no statistical adjustment for the effects of clustering. Finally, the study is less applicable to patients in the United States, in whom cirrhosis and thus HCC are usually secondary to hepatitis C, and the results probably have limited applicability to contemporary practice, in which the threshold for imaging for symptoms may be lower and the number of patients with incidentally discovered HCC on imaging is higher. The second trial used patient-level randomization stratified by township to assign patients with hepatitis B from 1989 to 1992 to the screening intervention (n= 3712), which consisted of serial -fetoprotein tests followed by ultrasonography for high -fetoprotein values, or the usual care group (n= 1869) (21). The population-based cancer registry used active case-finding techniques, and mortality was ascertained through the cancer registry and a population-based vital status registry. Cancer staging and cause of death were assessed by personnel blinded to intervention status. Only 28.8% of screening group participants completed all scheduled testing, but all participants completed at least 1 screening test. Fewer patients had stage III HCC in the screening group (19.8% vs. 41.0%; P value not reported). Hepatocellular carcinoma mortality was similar in both groups (1138 vs. 1114 per 100000 person-years; P= 0.86), as was all-cause mortality (1843 vs. 1788 per 100000 person-years; P value not s

“Did I Do as Best as the System Would Let Me?” Healthcare Professional Views on Hospital to Home Care Transitions

ABSTRACTBACKGROUNDPatients are vulnerable to poor quality, fragmented care as they transition from hospital to home. Few studies examine the discharge process from the perspectives of multiple healthcare professionals.OBJECTIVETo understand care transitions from the perspective of diverse healthcare professionals, and identify recommendations for process improvement.DESIGNCross sectional qualitative study.PARTICIPANTS AND SETTINGClinicians, care teams, and administrators from the inpatient general medicine services at one urban, academic hospital; two outpatient primary care clinics; and one Medicaid managed care plan.APPROACHWe conducted 13 focus groups and two in-depth interviews with participants prior to initiating a hospital-funded, multi-component transitional care intervention for uninsured and low-income publicly insured patients, the Care Transitions Innovation (C-TraIn). We used thematic analysis to identify emergent themes and a cross-case comparative analysis to describe variation by participant role and setting.KEY RESULTSPoor transitional care reflected healthcare system fragmentation, limiting the ability of healthcare professionals to provide optimal patient care. Lack of standardized processes, poor multidisciplinary communication within the hospital, and fragmented communication across settings led to chaotic, unsystematic transitions, poor patient outcomes, and feelings of futility and dissatisfaction among providers. Patients with complex psychosocial needs were especially vulnerable during care transitions. Recommended changes to improve transitional care included improving hospital multidisciplinary hospital rounds, clarifying accountability as patients move across settings, standardizing discharge processes, and providing additional medical staff training.CONCLUSIONSHospital to home care transitions are critical junctures that can impact health outcomes, experience of care, and costs. Transitional care quality improvement initiatives must address system fragmentation, reduce communication barriers within and between settings, and ensure adequate professional training.

The Effects of Pay-for-Performance Programs on Health, Health Care Use, and Processes of Care: A Systematic Review

Pay-for-performance (P4P) programs provide financial rewards or penalties to individual health care providers, groups of providers, or institutions according to their performance on measures of quality. In theory, if properly targeted and designed, P4P programs would help drive the behavior of providers and health care systems to improve the quality of care delivered, reduce unnecessary use of expensive health care services, and improve patient health outcomes (1). The idea is particularly relevant in the United States, where serious and broad gaps in health care quality have been tied in part to the long-standing fee-for-service system, which may provide incentives for service volume rather than quality (2). Despite their intuitive appeal, the promise of P4P programs in improving outcomes has not been empirically realized in past studies (36). The most recent systematic review examining the effectiveness of P4P programs in the United States found mixed evidence that P4P was associated with modest improvements in process-of-care outcomes but had little effect on patient outcomes (7). However, the literature has grown considerably since this review (which searched through 2012), and other countries, such as the United Kingdom, have gained considerable experience with large P4P initiatives that may provide information relevant to the United States. The purpose of the current review is to update and expand the prior systematic review in order to summarize current understanding of the effects of P4P programs targeted at physicians, groups, and institutions on process-of-care and patient outcomes in ambulatory and outpatient settings in and outside the United States. Methods This review was conducted according to a protocol that was developed using established reporting standards and posted to a public Web site (8) before the study was initiated (Appendix 1 of the Supplement). We used an analytic framework based on work by Damberg and colleagues (7) (Appendix 2 of the Supplement). Supplement. Supplemental Materials Data Sources and Searches We searched the following databases for studies that evaluated P4P programs: PubMed (1 June 2007 to 6 October 2016), MEDLINE, PsycINFO, CINAHL, Business Economics and Theory, Business Source Elite, Scopus, Faculty of 1000, and Gartner Research (1 June 2007 to 29 February 2016). We also performed targeted Google and PubMed searches aimed at well-known P4P demonstrations. We obtained additional articles from reference lists of pertinent studies, reviews, editorials, and expert recommendations. The search strategies are detailed in Appendix 3 of the Supplement. Study Selection Investigators reviewed titles and abstracts identified from literature searches. Two investigators independently assessed each potentially relevant article for inclusion using preestablished criteria (Appendices 4 and 5 of the Supplement). We included English-language studies of adult patients that evaluated ambulatory care or hospital-based P4P programs targeting health care providers at the individual, group, managerial, or institutional level and that reported any process-of-care, utilization, health, or intermediate health (clinical measures, such as a laboratory value or blood pressure) outcome. We included studies from other countries that have health systems similar to portions of the U.S. health care system. Studies examining only patient-targeted financial incentives, as well as payment models other than direct P4P, such as managed care, capitation, bundled payments, and accountable care organizations, were excluded. We also excluded studies that were not conducted in hospital or ambulatory settings, such as studies in long-term care facilities or nursing homes. We included clinical or cluster randomized, controlled trials (RCTs) of any size. We used a best-evidence approach, which is a method of specifying minimum inclusion criteria for nonrandomized studies (9). Inclusion of observational studies was limited to those with a comparison group, interrupted time series (ITS) studies, or large (n> 10000) cross-sectional or uncontrolled beforeafter studies. We excluded smaller uncontrolled studies because we had identified a large number of potentially relevant studies during a preliminary search and because the smaller uncontrolled studies were less likely to provide broadly applicable information given their limited scope and inherent methodological deficiencies. Data Extraction and Quality Assessment One investigator abstracted data elements from each included study, which were reviewed for accuracy by at least 1 additional investigator. We abstracted information on study design, sample size, country, program description, incentive structure (size and timing), target of the incentive, comparator, and outcomes (grouped as health, intermediate health, process-of-care, and utilization measures). Appendices 6 and 7 of the Supplement report these data. We classified studies according to 4 broad groupings: RCTs, ITS studies, controlled beforeafter studies, and uncontrolled beforeafter studies. Two investigators independently assessed study quality using the Newcastle-Ottawa Scale (10) for observational studies and the Cochrane Risk-of-Bias tool (11) for RCTs (Appendix 8 of the Supplement). Disagreements were resolved by consensus. Data Synthesis and Analysis We qualitatively synthesized the results of ambulatory and hospital studies separately and report process-of-care and patient outcomes for each setting. We synthesized results for specific P4P programs whenever possible. The review team evaluated the strength of the evidence according to guidance from the Agency for Healthcare Research and Quality (12). We did not perform meta-analysis because of the marked clinical heterogeneity across studies and the large number of observational studies. Role of the Funding Source The U.S. Department of Veterans Affairs Quality Enhancement Research Initiative supported this review but had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. Results Search Results We reviewed 3418 titles and abstracts, identified 586 potentially eligible full-text articles, and ultimately included 69 studies (Figure). Fifty-eight studies were in ambulatory settings (Table 1 and Appendix 6 of the Supplement), 11 were in hospital settings (Table 2 and Appendix 7 of the Supplement), 52 reported process-of-care outcomes, and 38 assessed patient outcomes. The studies examined a wide range of P4P programs with varying incentive structures, goals, and contexts. The programs also differed in their purposes and targets, but the largest number of studies focused on managing chronic conditions in the primary care setting. Studies were conducted in a wide range of countries, including the United Kingdom (27 studies), the United States (17 studies), Taiwan (13 studies), France (3 studies), the Netherlands (3 studies), Canada (3 studies), Australia (1 study), South Korea (1 study), and Italy (1 study). There were 2 RCTs and 67 observational studies (10 ITS studies, 37 controlled beforeafter studies, and 20 large uncontrolled beforeafter studies). Figure. Literature flow diagram. P4P= pay-for-performance. * The current systematic review updates and expands on the review by Damberg and colleagues (7). Table 1. Findings From Studies of Ambulatory-Based Pay-for-Performance Programs Table 2. Findings From Studies of Hospital-Based Pay-for-Performance Programs A large number of studies evaluated different aspects of 2 large-scale national programs: the United Kingdoms Quality and Outcomes Framework (QOF) (24 studies) and Taiwans diabetes mellitus (DM-P4P) program (9 studies). The QOF is a nationwide program that began in 2004. It incentivizes primary care practices to achieve quality indicators that support clinical care and public health goals. Incentive payments can comprise up to approximately 30% of total income. Practices are aided by integrated health information technology that delivers automated prompts and decision support (36, 83). Taiwans DM-P4P program, implemented in 2001, allows physicians to voluntarily enroll in the program, and they in turn are given freedom to choose which patients to enroll (51). From 2001 to 2006, incentives targeted process-of-care outcomes, which were augmented with intermediate health outcome measures after 2006. Ambulatory CareBased Programs Process-of-Care Outcomes We found 9 studies from the United States evaluating the effects of P4P on process-of-care outcomes (14, 1620, 2224). Most of these studies examined outcomes over 4 years and had an average follow-up of 2.5 years; very few studies reported longer-term data. One RCT found that individual incentives increased appropriate response to high blood pressure but not use of guideline-recommended antihypertensive medication (14). Of the 6 studies that reported positive results (16, 18, 19, 2224), 1 did not have a control group (24), and selection bias was a serious concern in 3 others because of the way the control group was chosen (18, 22, 23). Two methodologically sound controlled beforeafter studies found no improvements in processes of care (17, 20). In general, there was evidence across 17 studies in the United Kingdom (2631, 33, 3638, 4147) that the QOF was associated with improvements in process-of-care measures, although the evidence was mixed among the more methodologically rigorous studies. There were 6 ITS studies. One showed substantial improvements in the prescription of long-acting reversible contraceptives (26), and another showed modest improvement in the initiation of diabetes medications (27). Another study found increased rates of depression screening and diagnoses, but antidepressant prescribing remained unchanged (31). In the other 3 studies, improvements had begun well bef

Early Warning System Scores for Clinical Deterioration in Hospitalized Patients: A Systematic Review

RATIONALE Early warning system (EWS) scores are used by hospital care teams to recognize early signs of clinical deterioration and trigger more intensive care. OBJECTIVE To systematically review the evidence on the ability of early warning system scores to predict a patients risk of clinical deterioration and the impact of early warning system implementation on health outcomes and resource utilization. METHODS We searched the MEDLINE, CINAHL, and Cochrane Central Register of Controlled Trials databases through May 2014. We included English-language studies of early warning system scores used with adults admitted to medical or surgical wards. We abstracted study characteristics, including population, setting, sample size, duration, and criteria used for early warning system scoring. For predictive ability, the primary outcomes were modeled for discrimination on 48-hour mortality, cardiac arrest, or pulmonary arrest. Outcomes for the impact of early warning system implementation included 30-day mortality, cardiovascular events, use of vasopressors, respiratory failure, days on ventilator, and resource utilization. We assessed study quality using a modified Quality in Prognosis Studies assessment tool where applicable. MEASUREMENTS AND MAIN RESULTS Of 11,183 citations studies reviewed, one controlled trial and 20 observational studies of 13 unique models met our inclusion criteria. In eight studies, researchers addressed the predictive ability of early warning system tools and found a strong predictive value for death (area under the receiver operating characteristic curve [AUROC], 0.88-0.93) and cardiac arrest (AUROC, 0.74-0.86) within 48 hours. In 13 studies (one controlled trial and 12 pre-post observational studies), researchers addressed the impact on health outcomes and resource utilization and had mixed results. The one controlled trial was of good quality, and the researchers found no difference in mortality, transfers to the ICU, or length of hospital stay. The pre-post designs of the remaining studies have significant methodological limitations, resulting in insufficient evidence to draw conclusions. CONCLUSIONS Early warning system scores perform well for prediction of cardiac arrest and death within 48 hours, although the impact on health outcomes and resource utilization remains uncertain, owing to methodological limitations. Efforts to assess performance and effectiveness more rigorously will be needed as early warning system use becomes more widespread.

Treatment of anemia in patients with heart disease: A systematic review

BACKGROUND The benefits of anemia treatment in patients with heart disease are uncertain. PURPOSE To evaluate the benefits and harms of treatments for anemia in adults with heart disease. DATA SOURCES MEDLINE, EMBASE, and Cochrane databases; clinical trial registries; reference lists; and technical advisors. STUDY SELECTION English-language trials of blood transfusions, iron, or erythropoiesis-stimulating agents in adults with anemia and congestive heart failure or coronary heart disease and observational studies of transfusion. DATA EXTRACTION Data on study design, population characteristics, hemoglobin levels, and health outcomes were extracted. Trials were assessed for quality. DATA SYNTHESIS Low-strength evidence from 6 trials and 26 observational studies suggests that liberal transfusion protocols do not improve short-term mortality rates compared with less aggressive protocols (combined relative risk among trials, 0.94 [95% CI, 0.61 to 1.42]; I2 = 16.8%), although decreased mortality rates occurred in a small trial of patients with the acute coronary syndrome (1.8% vs. 13.0%; P = 0.032). Moderate-strength evidence from 3 trials of intravenous iron found improved short-term exercise tolerance and quality of life in patients with heart failure. Moderate- to high-strength evidence from 17 trials of erythropoiesis-stimulating agent therapy found they offered no consistent benefits, but their use may be associated with harms, such as venous thromboembolism. LIMITATIONS Few trials have examined transfusions in patients with heart disease, and observational studies are potentially confounded by indication. Data supporting iron use come mainly from 1 large trial, and long-term effects are unknown. CONCLUSION Higher transfusion thresholds do not consistently improve mortality rates, but large trials are needed. Intravenous iron may help to alleviate symptoms in patients with heart failure and iron deficiency and also warrants further study. Erythropoiesis-stimulating agents do not seem to benefit patients with mild to moderate anemia and heart disease and may be associated with serious harms. PRIMARY FUNDING SOURCE U.S. Department of Veterans Affairs.

The Effects of Cannabis Among Adults With Chronic Pain and an Overview of General Harms: A Systematic Review

The use of medicinal cannabis has become increasingly accepted in the United States and globally (1, 2). Eight states and the District of Columbia have legalized cannabis for recreational purposes, and 28 states and the District of Columbia have legalized it for medical purposes (3). Between 45% and 80% of persons who seek medical cannabis do so for pain management (4, 5). Among patients who are prescribed long-term opioid therapy for pain, up to 39% are also using cannabis (6, 7). Physicians will increasingly need to engage in evidence-based discussions with their patients about the potential benefits and harms of cannabis use. However, little comprehensive and critically appraised information exists about the benefits and harms of using cannabis to treat chronic pain. The objectives of this systematic review were to assess the efficacy of cannabis for treating chronic pain and to provide a broad overview of the short- and long-term physical and mental health effects of cannabis use in chronic pain and general patient populations. Methods Topic Development This article is part of a larger report commissioned by the Veterans Health Administration (8). A protocol describing the review plan was posted to a publicly accessible Web site before the study began (9). Data Sources and Searches We searched MEDLINE, Embase, PubMed, PsycINFO, Evidence-Based Medicine Reviews (including Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effectiveness, Health Technology Assessments, and Cochrane Central Register of Controlled Trials), and gray literature sources from database inception through February 2016. We updated this search specifically for new randomized controlled trials (RCTs) and systematic reviews in March 2017. We obtained additional articles from systematic reviews, reference lists, and expert recommendations. We also searched for ongoing, unpublished, or recently completed studies on ClinicalTrials.gov, the International Clinical Trials Registry Platform, the ISRCTN Registry, National Institutes of Health RePORTER, and the Agency for Healthcare Research and Quality Grants On-Line Database. Supplement 1 provides details on the search strategy, which we developed in consultation with a research librarian. Supplement. Online Supplement Study Selection We included English-language studies assessing the effect on nonpregnant adults of plant-based cannabis preparations or whole-plant extracts, such as nabiximols, a nonsynthetic pharmaceutical product with a standard composition and dose (oromucosal spray delivering tetrahydrocannabinol [THC], 2.7 mg, and cannabidiol, 2.5 mg). We did not include synthesized, pharmaceutically prepared cannabinoids, such as dronabinol or nabilone, because they are not available in dispensaries, and the efficacy of synthetic cannabinoid preparations for chronic pain was examined in 2 recent reviews (10, 11). We broadly defined plant-based cannabis preparations to include any preparation of the cannabis plant itself (for example, cannabis cigarettes and oils) or cannabis plant extracts to capture the variety of products available in U.S. dispensaries (12). To address the efficacy of cannabis for treating chronic pain, we included controlled clinical trials and cohort studies. This review focuses specifically on pain outcomes, although our larger report and search were designed to include other outcomes, such as sleep and quality of life (8). Because data about harms in the general population might be applicable to chronic pain populations, we examined harms broadly and reported whether the data were derived from studies of the general population or populations with chronic pain. To capture potential cannabis-related harms that may be of interest to clinicians and patients, but whose prevalence has not been well-characterized in larger-scale observational studies, we also included case series and descriptive studies of emerging harms. Supplement 2 provides the criteria we used for study selection. We searched for primary literature and systematic reviews; we dual-reviewed 5% of identified abstracts and all of the included full-text articles to ensure reliability. Disagreements were resolved by a third reviewer. Given the broad scope of this review, we summarized data from existing systematic reviews. We included only reviews that clearly reported their search strategy, reported inclusion and exclusion criteria, and appraised the internal validity of the included trials (13). We prioritized the most recent reviews and those with the broadest scope. In addition, we included individual studies that met inclusion criteria and either were published after the end search date of the included review or were not included in a prior systematic review. Data Extraction and Quality Assessment For all reports, 2 investigators abstracted details of study design, setting, patient population, intervention, and follow-up, as well as important co-interventions, health outcomes, health care use, and harms. Two reviewers independently assessed each trial (including those that were identified from a prior systematic review) as having low, high, or unclear risk of bias (ROB) for the pain outcome using a tool developed by the Cochrane Collaboration (14). Disagreements were resolved by consensus. To assess the ROB of observational studies for the pain outcome, we considered potential sources of bias most relevant to this evidence base and adapted existing assessment tools (15, 16) (Supplement 3). Data Synthesis and Analysis For the subgroup of neuropathic pain studies, we did a study-level meta-analysis of the proportion of patients experiencing clinically significant (30%) pain relief (Supplement 4), and we used the profile-likelihood random-effects model (17) to combine risk ratios. We assessed the magnitude of statistical heterogeneity among the studies using the standard Cochran chi-square test, the I 2 statistic (18). All analyses were done using Stata/IC, version 13.1 (StataCorp). Clinical heterogeneity, variation in outcomes reported, and the small number of trials precluded meta-analysis for other subgroups and outcomes, so we reported these qualitatively. After group discussion, we classified the overall strength of evidence for each outcome as high, moderate, low, or insufficient on the basis of the consistency, coherence, and applicability of the body of evidence as well as the internal validity of individual studies (19, 20). Role of the Funding Source The U.S. Department of Veterans Affairs Quality Enhancement Research Initiative supported the review but had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review, and approval of the manuscript; or decision to submit the manuscript for publication. Results After reviewing 13674 titles and abstracts, we included 13 systematic reviews and 62 primary studies (Figure). Table 1 provides study-level details and the ROB rating for each of the chronic pain trials. Table 2 summarizes findings, including the ROB rating, by pain subgroup. Table 3 summarizes the harms in both pain and general populations. Supplement 5 provides additional study-level data from pain studies not included in prior reviews and from studies on general harms. Figure. Literature flow diagram. * Includes Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, Health Technology Assessments, and Cochrane Central Register of Controlled Trials. Table 1. Characteristics and Findings of RCTs on Cannabis Extracts for Treating Chronic Pain* Table 2. Summary of Evidence of the Benefits of Cannabis in Populations With Chronic Pain Table 3. Summary of Evidence for the Harms of Cannabis in Chronic Pain and General Adult Populations Effects of Cannabis in Treating Chronic Pain We identified 22 RCTs (2142) from 2 recently published systematic reviews (10, 11) and an additional 8 studies (5 RCTs [4347] and 3 cohort studies [4850]) that met our inclusion criteria and were not included in prior reviews. The primary methods of continuous pain assessment were a visual analogue scale from 0 to 100 mm and a numerical rating scale (NRS) from 0 to 10 (where 0 indicated no pain and 10 indicated the worst possible pain). Some of the studies identified the proportion of participants who had clinically significant improvements in pain intensity (defined as 30% reduction, or approximately 2 points on the NRS and 20 mm on the visual analogue scale). Neuropathic Pain Thirteen trials examined the effects of cannabis-based preparations on neuropathic pain (Table 1). Participants had central or peripheral neuropathic pain related to various health conditions. Of these studies, 11 were rated as having low ROB (24, 27, 28, 30, 31, 33, 36, 39, 40, 43, 47), 1 as having unclear ROB (26), and 1 as having high ROB (35). Overall, we found low-strength evidence that cannabis may alleviate neuropathic pain in some patients (Table 2). Studies generally did not find clinically significant between-group differences on continuous pain scales, but a higher proportion of intervention patients had clinically significant pain relief up to several months later. Across 9 studies, intervention patients were more likely to report at least 30% improvement in pain (risk ratio, 1.43 [95% CI, 1.16 to 1.88]; I 2= 38.6%; P= 0.111) (Supplement 4). Most studies were small, few reported outcomes beyond 2 to 3 weeks, and none reported long-term outcomes. In the largest RCT, 246 patients with peripheral neuropathic pain self-titrated nabiximols up to a maximum dosage of 24 sprays per day or received a placebo (27). Those who completed the study (79 in the nabiximols group and 94 in the placebo group) and responded positively to the intervention had a significant decrease in pain (odds ratio, 1.97 [CI, 1.05 to 3.70]). However, among all participants, including those who did not have an intervention response, the reductio