Avishek Nagi

Evidence-Based Risk Communication: A Systematic Review

Shared decision making is a collaborative process that allows patients and medical professionals to consider the best scientific evidence available, along with patients values and preferences, to make health care decisions (1). A recent Institute of Medicine report concluded that although people desire a patient experience that includes deep engagement in shared decision making, there are gaps between what patients want and what they get (2). For patients to get the experience they want, providers must effectively communicate evidence about benefits and harms. To improve the decision-making process, the Institute of Medicine recommended development and dissemination of high-quality communication tools (2). New tools, however, must match patients numerical abilities, which are often limited. For example, in one study, as many as 40% of high school graduates could not perform basic numerical operations, such as converting 1% of 1000 to 10 of 1000. This collective statistical illiteracy is a major barrier to the interpretation of health statistics (3). Physicians may also find statistical information difficult to interpret and explain (4). Existing literature about methods of communicating benefits and harms is broad. One review, based on 19 studies, concluded that the choice of a specific graphic is not as important as whether the graphic frames the frequency of an event with a visual representation of the total population in which it occurs (5). Another review, involving a limited literature search, found that comprehension improved when using frequencies (such as 1 in 5) instead of event rates (such as 20%) and using absolute risk reductions (ARRs) instead of relative risk reductions (RRRs) (6). The review did not assess affective outcomes, such as patient satisfaction, and behavioral outcomes, such as changes in decision making. Yet another review identified strong evidence that patients misinterpret RRRs and supported the effectiveness of graphs in communicating harms (7). However, they did not examine the comparative effectiveness of such approaches. More narrowly focused Cochrane reviews examined the communication of risk specific to screening tests (8, 9); numerical presentations, such as ARRs, RRRs, and numbers needed to treat (NNTs) (10); and effects of decision aids (11). An expert commentary about effective risk communication recommended using plain language, icon arrays, and absolute risks and providing time intervals with risk information (12). A group of experts identified 11 key components of risk communication, including presenting numerical estimates in context with evaluative labels, conveying uncertainty, and tailoring estimates (13). The aim of this systematic review is to comprehensively examine the comparative effectiveness of all methods of communicating probabilistic information about benefits and harms to patients to maximize their understanding, satisfaction, and decision-making ability. Methods We developed and followed a plan for the review that included several searches and dual abstraction of study data using standardized abstraction forms. Data Sources and Study Selection We searched PubMed (1966 to March 2014), CINAHL, EMBASE, and the Cochrane Central Register of Controlled Trials (1966 to December 2011) using keywords and structured terms related to the concepts of patients; communication; riskbenefit; and outcomes, such as understanding or comprehension, preferences or satisfaction, and decision making. Supplement 1 shows the detailed search strategy. Supplement 1. Search Strategies We included cross-sectional or prospective, longitudinal trials that were published in English and had an active control group that recruited patients or healthy volunteers and compared any method of communicating probabilistic information with another method. We focused on different methods of communicating the same specific probabilities to eliminate any independent effects that could result from different probabilities being studied (for example, different magnitudes or directions of effect). Studies of personalized risks, which may vary from person to person, were included when participants were randomly assigned. When studies of personalized risks were not randomized, the risks were considered to differ between the groups and were excluded. No limits were placed on study size, location, or duration or on the nature of the communication method. When needed, we reviewed sources specified in the articles, such as Web sites, to directly review the interventions and determine whether probabilistic information was addressed. Studies of medical students, health professionals, and public health or mass media campaigns were excluded. One independent reviewer screened each title and abstract and excluded citations that were not original studies or were unrelated to probabilistic information. Two independent reviewers screened the full text of the remaining citations to identify eligible articles. Disagreements between the 2 reviewers were resolved by consensus, with a third reviewer arbitrating any unresolved disagreements. Data Extraction and Quality Assessment Two reviewers independently abstracted detailed information about the study population, interventions, primary outcomes, and risk of bias from each included study using a standardized abstraction form, which was developed a priori (Supplement 2). A third reviewer resolved any disagreements. We categorized outcomes in 1 of 3 domains: cognitive (or understanding, such as accuracy in answering questions related to probabilistic information, or general comprehension of the probabilistic information), affective (such as preferences for or satisfaction with the method of communicating probabilistic information), and behavioral (such as real or theoretical decision making). Supplement 2. Abstraction Form Risk of bias in randomized, controlled trials was assessed on the basis of adequacy of randomization, allocation concealment, similarity of study groups at baseline, blinding, equal treatment of groups throughout the study, completeness of follow-up, and intention to treat (participants analyzed in the groups to which they were randomly assigned) (14). Risk of bias in observational studies was assessed with a modified set of criteria adapted from the NewcastleOttawa Scale (15). Data Synthesis and Analysis Data were tabulated, and the frequency of all head-to-head comparisons in studies was assessed to identify clusters of comparisons. In many instances, several interventions were bundled in a single study group (such as event rate plus icon array, or event rate plus natural frequencies plus ARRs). Bundles were not separated or combined with similar interventions because it could not be determined which component of the bundle drove the intervention. Descriptive statistics were used. We decided a priori not to do meta-analysis because of study heterogeneity. We emphasized findings from randomized studies as well as nonrandomized studies when findings were supported by more than 1 study. Role of the Funding Source No funding supported this study. The authors participated within their role on the Evidence-Based Medicine Task Force of the Society of General Internal Medicine. Results The initial search through December 2011 retrieved 22103 citations (16661 from PubMed, 1194 from CINAHL, 2861 from the Cochrane Central Register of Controlled Trials, and 1387 from EMBASE), and 20076 remained after removing duplicates. We updated the PubMed search through 30 March 2014, yielding 6529 additional citations; 5970 remained after removing duplicates, for a total of 26046 citations for review. A total of 630 articles were selected for full-text review and 84 were included, representing 91 unique studies (1699). Reasons for exclusion are noted in Figure 1, and study details are provided in Supplement 3. Figure 1. Summary of evidence search and selection. Supplement 3. Details of All Included Studies Seventy-four (81.3%) of the 91 included studies were randomized trials, most with cross-sectional designs. The median number of participants in randomized trials was 268 (range, 31 to 4685), and the median in all studies was 268 (range, 24 to 16133). Thirty-three studies (36.3%) included patients at specific risk for the target condition of interest. Forty-eight studies (52.7%) presented probabilistic data about benefits of a therapy or intervention (with 7 [14.6%] also presenting harms), 21 (23.1%) presented data only on harms, and 9 (10%) involved screening tests. Forty-nine studies (54.4%) delivered interventions on paper and 39 (42.9%) on a computer, typically over the Internet. The characteristics of study participants are presented in Tables 1 and 2. Table 1. Characteristics of Study Participants Table 2. Proportion of Studies Including Participants at Risk Versus Not at Risk for Target Condition Risk of bias for the included randomized trials was moderate (Figure 2). Randomization was adequate in 32 trials (42.7%), inadequate in 3 (4.0%), and unclear in 40 (53.3%). Allocation concealment was not stated in 55 trials (73.3%). Similarity of groups at baseline was adequate in 37 trials (49.3%) and unclear in 32 (42.7%). Blinding, equal treatment, and intention-to-treat items were similarly difficult to assess from reported information. Figure 2. Risk of bias for randomized, controlled trials (n = 74). Adapted from reference 100. Study Interventions and Comparators A frequency table (heat map) of all study intervention comparisons was created to identify clusters of comparisons (Supplement 4). The heat map represents study group comparisons, so one study may contribute several comparisons. The most commonly studied numerical presentations of data were natural frequencies, defined as the numbers of persons with events juxtaposed with a baseline denominator of persons (for example, 4 out of 100 persons had the outcome); event rates, defined as the proportions of persons wi

Evidence-based risk communication

Shared decision making is a collaborative process that allows patients and medical professionals to consider the best scientific evidence available, along with patients values and preferences, to make health care decisions (1). A recent Institute of Medicine report concluded that although people desire a patient experience that includes deep engagement in shared decision making, there are gaps between what patients want and what they get (2). For patients to get the experience they want, providers must effectively communicate evidence about benefits and harms. To improve the decision-making process, the Institute of Medicine recommended development and dissemination of high-quality communication tools (2). New tools, however, must match patients numerical abilities, which are often limited. For example, in one study, as many as 40% of high school graduates could not perform basic numerical operations, such as converting 1% of 1000 to 10 of 1000. This collective statistical illiteracy is a major barrier to the interpretation of health statistics (3). Physicians may also find statistical information difficult to interpret and explain (4). Existing literature about methods of communicating benefits and harms is broad. One review, based on 19 studies, concluded that the choice of a specific graphic is not as important as whether the graphic frames the frequency of an event with a visual representation of the total population in which it occurs (5). Another review, involving a limited literature search, found that comprehension improved when using frequencies (such as 1 in 5) instead of event rates (such as 20%) and using absolute risk reductions (ARRs) instead of relative risk reductions (RRRs) (6). The review did not assess affective outcomes, such as patient satisfaction, and behavioral outcomes, such as changes in decision making. Yet another review identified strong evidence that patients misinterpret RRRs and supported the effectiveness of graphs in communicating harms (7). However, they did not examine the comparative effectiveness of such approaches. More narrowly focused Cochrane reviews examined the communication of risk specific to screening tests (8, 9); numerical presentations, such as ARRs, RRRs, and numbers needed to treat (NNTs) (10); and effects of decision aids (11). An expert commentary about effective risk communication recommended using plain language, icon arrays, and absolute risks and providing time intervals with risk information (12). A group of experts identified 11 key components of risk communication, including presenting numerical estimates in context with evaluative labels, conveying uncertainty, and tailoring estimates (13). The aim of this systematic review is to comprehensively examine the comparative effectiveness of all methods of communicating probabilistic information about benefits and harms to patients to maximize their understanding, satisfaction, and decision-making ability. Methods We developed and followed a plan for the review that included several searches and dual abstraction of study data using standardized abstraction forms. Data Sources and Study Selection We searched PubMed (1966 to March 2014), CINAHL, EMBASE, and the Cochrane Central Register of Controlled Trials (1966 to December 2011) using keywords and structured terms related to the concepts of patients; communication; riskbenefit; and outcomes, such as understanding or comprehension, preferences or satisfaction, and decision making. Supplement 1 shows the detailed search strategy. Supplement 1. Search Strategies We included cross-sectional or prospective, longitudinal trials that were published in English and had an active control group that recruited patients or healthy volunteers and compared any method of communicating probabilistic information with another method. We focused on different methods of communicating the same specific probabilities to eliminate any independent effects that could result from different probabilities being studied (for example, different magnitudes or directions of effect). Studies of personalized risks, which may vary from person to person, were included when participants were randomly assigned. When studies of personalized risks were not randomized, the risks were considered to differ between the groups and were excluded. No limits were placed on study size, location, or duration or on the nature of the communication method. When needed, we reviewed sources specified in the articles, such as Web sites, to directly review the interventions and determine whether probabilistic information was addressed. Studies of medical students, health professionals, and public health or mass media campaigns were excluded. One independent reviewer screened each title and abstract and excluded citations that were not original studies or were unrelated to probabilistic information. Two independent reviewers screened the full text of the remaining citations to identify eligible articles. Disagreements between the 2 reviewers were resolved by consensus, with a third reviewer arbitrating any unresolved disagreements. Data Extraction and Quality Assessment Two reviewers independently abstracted detailed information about the study population, interventions, primary outcomes, and risk of bias from each included study using a standardized abstraction form, which was developed a priori (Supplement 2). A third reviewer resolved any disagreements. We categorized outcomes in 1 of 3 domains: cognitive (or understanding, such as accuracy in answering questions related to probabilistic information, or general comprehension of the probabilistic information), affective (such as preferences for or satisfaction with the method of communicating probabilistic information), and behavioral (such as real or theoretical decision making). Supplement 2. Abstraction Form Risk of bias in randomized, controlled trials was assessed on the basis of adequacy of randomization, allocation concealment, similarity of study groups at baseline, blinding, equal treatment of groups throughout the study, completeness of follow-up, and intention to treat (participants analyzed in the groups to which they were randomly assigned) (14). Risk of bias in observational studies was assessed with a modified set of criteria adapted from the NewcastleOttawa Scale (15). Data Synthesis and Analysis Data were tabulated, and the frequency of all head-to-head comparisons in studies was assessed to identify clusters of comparisons. In many instances, several interventions were bundled in a single study group (such as event rate plus icon array, or event rate plus natural frequencies plus ARRs). Bundles were not separated or combined with similar interventions because it could not be determined which component of the bundle drove the intervention. Descriptive statistics were used. We decided a priori not to do meta-analysis because of study heterogeneity. We emphasized findings from randomized studies as well as nonrandomized studies when findings were supported by more than 1 study. Role of the Funding Source No funding supported this study. The authors participated within their role on the Evidence-Based Medicine Task Force of the Society of General Internal Medicine. Results The initial search through December 2011 retrieved 22103 citations (16661 from PubMed, 1194 from CINAHL, 2861 from the Cochrane Central Register of Controlled Trials, and 1387 from EMBASE), and 20076 remained after removing duplicates. We updated the PubMed search through 30 March 2014, yielding 6529 additional citations; 5970 remained after removing duplicates, for a total of 26046 citations for review. A total of 630 articles were selected for full-text review and 84 were included, representing 91 unique studies (1699). Reasons for exclusion are noted in Figure 1, and study details are provided in Supplement 3. Figure 1. Summary of evidence search and selection. Supplement 3. Details of All Included Studies Seventy-four (81.3%) of the 91 included studies were randomized trials, most with cross-sectional designs. The median number of participants in randomized trials was 268 (range, 31 to 4685), and the median in all studies was 268 (range, 24 to 16133). Thirty-three studies (36.3%) included patients at specific risk for the target condition of interest. Forty-eight studies (52.7%) presented probabilistic data about benefits of a therapy or intervention (with 7 [14.6%] also presenting harms), 21 (23.1%) presented data only on harms, and 9 (10%) involved screening tests. Forty-nine studies (54.4%) delivered interventions on paper and 39 (42.9%) on a computer, typically over the Internet. The characteristics of study participants are presented in Tables 1 and 2. Table 1. Characteristics of Study Participants Table 2. Proportion of Studies Including Participants at Risk Versus Not at Risk for Target Condition Risk of bias for the included randomized trials was moderate (Figure 2). Randomization was adequate in 32 trials (42.7%), inadequate in 3 (4.0%), and unclear in 40 (53.3%). Allocation concealment was not stated in 55 trials (73.3%). Similarity of groups at baseline was adequate in 37 trials (49.3%) and unclear in 32 (42.7%). Blinding, equal treatment, and intention-to-treat items were similarly difficult to assess from reported information. Figure 2. Risk of bias for randomized, controlled trials (n = 74). Adapted from reference 100. Study Interventions and Comparators A frequency table (heat map) of all study intervention comparisons was created to identify clusters of comparisons (Supplement 4). The heat map represents study group comparisons, so one study may contribute several comparisons. The most commonly studied numerical presentations of data were natural frequencies, defined as the numbers of persons with events juxtaposed with a baseline denominator of persons (for example, 4 out of 100 persons had the outcome); event rates, defined as the proportions of persons wi

Effects of Nurse-Managed Protocols in the Outpatient Management of Adults With Chronic Conditions: A Systematic Review and Meta-analysis

Medical management of chronic illness consumes 75% of every health care dollar spent in the United States (1). Thus, provision of economical and accessibleyet high-qualitycare is a major concern. Diabetes mellitus, hypertension, and hyperlipidemia are prime examples of chronic diseases that cause substantial morbidity and mortality (2, 3) and require long-term medical management. For each of these disorders, most care occurs in outpatient settings where well-established clinical practice guidelines are available (47). Despite the availability of these guidelines, there are important gaps between the care recommended and the care delivered (810). The shortage of primary care clinicians has been identified as 1 barrier to the provision of comprehensive care for chronic disease (11, 12) and is an impetus to develop strategies for expanding the roles and responsibilities of other interdisciplinary team members to help meet this increasing need. The patient-centered medical home concept was developed in an effort to serve more persons and improve chronic disease care. It is a model of primary care transformation that builds on other efforts, such as the chronic care model (13), and includes the following elements: patient-centered orientation toward the whole person, team-based care coordinated across the health care system and community, enhanced access to care, and a systems-based approach to quality and safety. Care teams may include nurses, primary care providers, pharmacists, and behavioral health specialists. An organizing principle for care teams is to utilize personnel at the highest level of their skill set, which is particularly relevant given the expected increase in demand for primary care services resulting from the Patient Protection and Affordable Care Act. With this increased demand, the largest health care workforce, registered nurses (RNs), may be a valuable asset alongside other nonphysician clinicians, including physician assistants, nurse practitioners, and clinical pharmacists, to serve more persons and improve chronic disease care. Robust evidence supports the effectiveness of nurses in providing patient education about chronic disease and secondary prevention strategies (1419). With clearly defined protocols and training, nurses may also be able to order relevant diagnostic tests, adjust routine medications, and appropriately refer patients. Our purpose was to synthesize the current literature describing the effects of nurse-managed protocols, including medication adjustment, for the outpatient management of adults with common chronic conditions, namely diabetes, hypertension, and hyperlipidemia. Methods We followed a standard protocol for all steps of this review. A technical report that fully details our methods and presents results for all original research questions is available at www.hsrd.research.va.gov/publications/esp/reports.cfm. Data Sources and Searches In consultation with a master librarian, we searched MEDLINE (via PubMed), Cochrane Central Register of Controlled Trials, EMBASE, and CINAHL from 1 January 1980 through 31 January 2014 for English-language, peer-reviewed publications evaluating interventions that compared nurse-managed protocols with usual care in studies targeting adults with chronic conditions (Supplement 1). Supplement 1. Search Strategy We selected exemplary articles and used a Medical Subject Heading analyzer to identify terms for nurse protocols. We added selected free-text terms and validated search terms for randomized, controlled trials (RCTs) and quasi-experimental studies, and we searched bibliographies of exemplary studies and applicable systematic reviews for missed publications (15, 17, 2029). To assess for publication bias, we searched ClinicalTrials.gov to identify completed but unpublished studies meeting our eligibility criteria. Study Selection, Data Extraction, and Quality Assessment Two reviewers used prespecified eligibility criteria to assess all titles and abstracts (Supplement 2). Eligibility criteria included the involvement of an RN or a licensed practical nurse (LPN) functioning beyond the usual scope of practice, such as adjusting medications and conducting interventions based on a written protocol. Potentially eligible articles were retrieved for further evaluation. Disagreements on inclusion or exclusion were resolved by discussion or a third reviewer. Studies excluded at full-text review are listed in Supplement 3. Abstraction and quality assessment were done by 1 reviewer and confirmed by a second. We piloted the abstraction forms, designed specifically for this review, on a sample of included articles. Key characteristics abstracted included patient descriptors, setting, features of the intervention and comparator, match between the sample and target populations, extent of the nurse interventionists training, outcomes, and quality elements. Supplements 4 and 5 summarize quality criteria and ratings, respectively. Supplement 2. Eligibility Criteria Supplement 3. List of Excluded Studies Supplement 4. Criteria Used in Risk of Bias Assessment Supplement 5. Detailed Study Characteristics Because many studies were done outside the United States, we queried the authors of such studies about the education and scope of practice of the nurse interventionists. Authors were e-mailed a table detailing the credentialing and scope of practice of various U.S. nurses and asked to classify their nurse interventionist. Data Synthesis and Analysis The primary outcomes were the effects of nurse-managed protocols on biophysical markers (for example, glycosylated hemoglobin or hemoglobin A1c [HbA1c]), patient treatment adherence, nurse protocol adherence, adverse effects, and resource use. When quantitative synthesis (that is, meta-analysis) was feasible, dichotomous outcomes were combined using odds ratios and continuous outcomes were combined using mean differences in random-effects models. For studies with unique but conceptually similar outcomes, such as ordering a guideline-indicated laboratory test, we synthesized outcomes across conditions if intervention effects were sufficiently homogeneous. We used the Knapp and Hartung method (30, 31) to adjust the SEs of the estimated coefficients. For categories with several potential outcomes (for example, biophysical markers) that may vary across chronic conditions, we selected outcomes for each chronic condition a priori: HbA1c level for diabetes, blood pressure (BP) for hypertension, and cholesterol level for hyperlipidemia. In 1 example (32), we imputed missing SDs using estimates from similar studies. We computed summary estimates of effect and evaluated statistical heterogeneity using the Cochran Q and I 2 statistics. We did subgroup analyses to examine potential sources of heterogeneity, including where the study was conducted and intervention content. Subgroup analyses involved indirect comparisons and were subject to confounding; thus, results were interpreted cautiously. Publication bias was assessed using a ClinicalTrials.gov search and funnel plots when at least 10 studies were included in the analysis. When quantitative synthesis was not feasible, we analyzed data qualitatively. We gave more weight to evidence from higher-quality studies with more precise estimates of effect. The qualitative syntheses identified and documented patterns in efficacy and safety of the intervention across conditions and outcome categories. We analyzed potential reasons for inconsistency in treatment effects across studies by evaluating variables, such as differences in study population, intervention, comparator, and outcome definitions. We followed the approach recommended by the Agency for Healthcare Research and Quality (33) to evaluate the overall strength of the body of evidence. This approach assesses the following 4 domains: risk of bias, consistency, directness, and precision. These domains were considered qualitatively, and a summary rating of high, moderate, low, or insufficient evidence was assigned. Role of the Funding Source The Veterans Affairs Quality Enhancement Research Initiative funded the research but did not participate in the conduct of the study or the decision to submit the manuscript for publication. Results Our electronic and manual searches identified 2954 unique citations (Figure 1). Of the 23 potentially eligible studies, 4 were excluded because we could not verify whether nurses had the authority to initiate or titrate medications and the author did not respond to our query for clarification (3437). We excluded a trial of older adults in which we could not differentiate the target illnesses (38). Approximately two thirds of the authors we contacted for missing data or clarification responded. Figure 1. Summary of evidence search and selection. * Methods or follow-up articles. We included 18 unique studies (23004 patients) that focused on patients with elevated cardiovascular risk (Table) (32, 3955). Of these, 16 were RCTs and 2 were controlled before-and-after studies on diabetes (49, 53). The comparator was usual care in all but 1 study, in which a reverse-control design was used, and each intervention served as the control for the other. Eleven studies were done in Western Europe and 7 in the United States. Median age of participants was 58.3 years (range, 37.2 to 72.1 years) based on 16 studies. Approximately 47% of the participants were female. Race was not reported in 84% of the studies. Supplement 5 gives detailed study characteristics. No outstanding studies were identified through ClinicalTrials.gov. Supplement 6 provides funnel plots that assess publication bias. Table. Study and Patient Characteristics of Included Diabetes, Hypertension, and Hyperlipidemia Studies Supplement 6. Assessment of Publication Bias: Funnel Plots Overall, these studies displayed moderate risk of bias. Two studies were judged as having a high risk of bias because of inadequate randomization (44, 5

Clinical outcomes of metformin use in populations with chronic kidney disease, congestive heart failure, or chronic liver disease: A systematic review

After its approval by the U.S. Food and Drug Administration (FDA) in 1994, metformin became the recommended initial treatment for type 2 diabetes mellitus in the United States (1). Beyond its glycemic benefits, metformin typically does not cause weight gain or hypoglycemia and may be associated with lower mortality (2, 3). Because of concerns regarding lactic acidosis with use of phenformin, a related biguanide withdrawn from the market in 1977, the FDA applied a boxed warning to metformin concurrent with its approval (4). This warning cautioned against using metformin in the setting of chronic kidney disease (CKD), which may impair excretion of the drug, and recommended caution in patients with conditions that may promote lactate accumulation (such as congestive heart failure [CHF] and chronic liver disease [CLD]) (5). Despite this warning, recent estimates suggest that 20% to 30% of persons receiving metformin have historical contraindications or precautions regarding its use (6, 7). These findings reflect that many prescribers believe the FDA boxed warning is too restrictive (8, 9). Literature reviews indicate no clear association between metformin use and lactic acidosis (10) and suggest that the drug is safe for patients with moderate CKD or CHF (11, 12). In 2006, the FDA removed CHF as a contraindication to metformin use, although acute or unstable CHF remains a precaution (13, 14). In April 2016, the FDA revised its warning regarding metformin use in patients with CKD, switching from a serum creatininebased definition of renal impairment to more-inclusive criteria based on estimated glomerular filtration rate (eGFR) (15). With this change, an estimated 1 million additional patients with moderate CKD (eGFR, 30 to <60 mL/min/1.73 m2) became eligible to receive metformin, although severe CKD (eGFR, <30 mL/min/1.73 m2) remains a contraindication (16). In the wake of these changes, metformin use will continue to increase in populations with historical contraindications and precautions. Prescribers therefore must fully understand the consequences of metformin use in these groups. To promote informed prescribing, we systematically reviewed the literature regarding the benefits and harms of metformin use (beyond lactic acidosis) among patients with common chronic diseases historically identified by the FDAs boxed warning as contraindications or precautions: moderate to severe CKD, CHF, and CLD with impaired hepatic function. Methods Study Design This work was part of a Veterans Health Administration (VHA)-funded report. Additional details are available online (www.hsrd.research.va.gov/publications/esp). The present analysis focuses on the following question: For patients with type 2 diabetes and a historical contraindication or precaution regarding metformin use, what are the benefits and harms (beyond lactic acidosis) of metformin treatment? This review followed a published protocol (PROSPERO: CRD42016027708), and each step was pilot-tested to train and calibrate investigators. Data Sources and Study Selection In consultation with an expert medical librarian, we searched PubMed, the Cochrane Central Register of Controlled Trials, EMBASE, and the International Pharmaceutical Abstracts in November 2015; we subsequently updated our PubMed search through September 2016. We also searched ClinicalTrials.gov for relevant completed and ongoing studies. Appendix Table 1 presents our search strategies. We also screened reference lists of published reviews and queried Bristol-Myers Squibb, the manufacturer of the branded metformin formulation, for other relevant studies. Appendix Table 1. Search Strategies for Online Databases, With Date of Search and Specific Terms Our prespecified inclusion and exclusion criteria are listed in Appendix Table 2. We included English-language clinical trials and observational cohort studies that examined adults with type 2 diabetes and a metformin contraindication or precaution of interest (moderate to severe CKD [eGFR, <60 mL/min/1.73 m2], CHF, or CLD with hepatic impairment); compared antihyperglycemic regimens that included metformin with those that did not; and reported all-cause mortality, major adverse cardiovascular events (MACEs), glycemic control, lipid control, hypoglycemia, weight gain, or vitamin B12 deficiency. Our VHA stakeholders and technical expert panel provided guidance on outcome selection. Appendix Table 2. Inclusion and Exclusion Criteria for Citations Data Extraction and Quality Assessment of Individual Studies Two investigators screened all citations for eligibility, and citations considered relevant by either individual advanced to full-text review. Two investigators reviewed all full-text articles and resolved disagreements through discussion or adjudication by a third investigator. Before excluding any potentially eligible study whose primary analysis did not explicitly address a population with a metformin contraindication or precaution, we examined the full text for relevant subgroup analyses. Two investigators independently assessed study quality, and disagreements were resolved by consensus or through arbitration by a third investigator. Using published quality criteria, we developed a customized risk-of-bias (ROB) assessment tool designed to address selection, performance, attrition, detection, and reporting biases (Appendix Figure 1) (17). We assigned each study an ROB score (low, moderate, or high). Appendix Figure 1. Quality assessment for observational studies. BNP = brain natriuretic peptide; CHF = congestive heart failure; CKD = chronic kidney disease; CT = computed tomography; CV = cardiovascular; DM = diabetes mellitus; DM2 = type 2 diabetes mellitus; eGFR = estimated glomerular filtration rate; FBS = fasting blood sugar; f/u = follow-up; HbA1c = hemoglobin A1c; HTN = hypertension; ICD = International Classification of Diseases; MACEs = major adverse cardiovascular events; MI = myocardial infarction; NA = not applicable; NR = not reported; PE = physical examination. Appendix Figure 1. Continued Data Abstraction For each included study, an investigator abstracted data by using a customized DistillerSR database (Evidence Partners); a second investigator independently reviewed these data for accuracy. Relevant data included demographics, study setting, contraindication or precaution definitions, metformin dosage, other antihyperglycemic agents, comparators, and outcomes. We treated multiple publications from a single study as a single data point, prioritizing the longest-term and most complete results. If critical data were missing or unclear in a published report, we contacted the manuscript authors. Data Synthesis We developed summary tables to characterize all included studies for each metformin contraindication or precaution of interest. Of note, 2 studies (18, 19) separately compared distinct groups of metformin usersthose receiving metformin monotherapy and those receiving metforminsulfonylurea combination therapywith patients receiving sulfonylurea monotherapy. In each case, we derived a pooled, weighted hazard ratio (HR) for all metformin users, incorporating an approximation of the correlation resulting from the shared sulfonylurea monotherapy reference group (Appendix). For another study (20), we estimated the HR and variance from the reported frequencies and the odds ratio (OR) by using an established approach (21, 22) (Appendix). If 3 or more studies were conceptually similar in terms of design, population, intervention, and outcomes, we performed quantitative synthesis by using a random-effects model to generate summary HRs. For analyses with fewer than 20 studies, we used the KnappHartung approach to adjust the SEs of the estimated coefficients (23, 24). If appropriate, we conducted sensitivity analyses by omitting subgroups with more severe contraindications or precautions (such as an eGFR <30 mL/min/1.73 m2), studies with shorter follow-up (<2 years), and studies not using propensity-score adjustment. We evaluated statistical heterogeneity by using Cochran Q and I 2 statistics; for analyses including 10 or more studies, we assessed publication bias by using funnel plots and Begg and Egger tests (25, 26). For cases with too few studies to warrant meta-analysis, we performed qualitative synthesis. We conducted all quantitative analyses by using R (version 3.1.2) (The R Foundation), including the metafor package (version 1.9-7), for meta-analysis. Strength of Evidence We used the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach to evaluate the overall strength of evidence (SOE) for outcomes with sufficient data. Using the domains of ROB, directness, and consistency or precision of treatment effects, an investigator (J.W.W.) rated the SOE as high, moderate, low, or insufficient. We considered the effect of residual confounders, magnitude of effect, and publication bias (27, 28). Role of the Funding Source This review was funded by the U.S. Department of Veterans Affairs. The funding source had no role in the study design, data collection, analysis, preparation of the manuscript, or decision to submit the manuscript for publication. Results From 4910 screened citations, we reviewed 532 full-text articles and identified 17 eligible studies (Figure 1). All were observational and addressed populations with moderate to severe CKD (n= 5), CHF (n= 11), or CLD with hepatic impairment (n= 3); 3 studies addressed both CKD and CHF. Appendix Table 3 provides details on the included studies. Of note, we identified no ongoing studies in ClinicalTrials.gov that met our inclusion criteria. Figure 1. Flow of articles through the literature search and screening process. CHF = congestive heart failure; CKD = chronic kidney disease; CLD = chronic liver disease; OECD = Organization for Economic Cooperation and Development. * Search results are from EMBASE (n = 2512), PubMed (n = 2381), and Cochrane (n = 17). Three references

An evidence synthesis of care models to improve general medical outcomes for individuals with serious mental illness: A systematic review

OBJECTIVE To conduct a systematic review of studies of interventions that integrated medical and mental health care to improve general medical outcomes in individuals with serious mental illness. DATA SOURCES English-language publications in MEDLINE (via PubMed), EMBASE, PsycINFO, and the Cochrane Library, from database inception through January 18, 2013, were searched using terms for our diagnoses of interest, a broad set of terms for care models, and a set of terms for randomized controlled trials (RCTs) or quasi-experimental design. Bibliographies of included articles were examined for additional sources. ClinicalTrials.gov was searched using the terms for our diagnoses of interest (serious mental illness,SMI,bipolar disorder,schizophrenia,orschizoaffective disorder) to assess for evidence of publication bias and ongoing studies. STUDY SELECTION 4 RCTs were included from 1,729 articles reviewed. Inclusion criteria were RCT or quasi-experimental design; adult outpatient population with 25% or greater carrying a diagnosis of schizophrenia, schizoaffective disorder, or bipolar disorder; intervention with a stated goal to improve medical outcomes through integration of care, using a comparator of usual care or other quality improvement strategy; and outcomes assessing process of care, clinical outcomes, or physical functioning. DATA EXTRACTION A trained researcher abstracted the following data from the included articles: study design, funding source, setting, population characteristics, eligibility and exclusion criteria, number of subjects and providers, intervention(s), comparison(s), length of follow-up, and outcome(s). These abstracted data were then overread by a second reviewer. RESULTS Of the 4 studies reviewed, 2 good-quality studies (according to the guidelines of the Agency for Healthcare Research and Quality) that evaluated processes of preventive and chronic disease care demonstrated positive effects of integrated care. Specifically, integrated care interventions were associated with increased rates of immunization and screening. All 4 RCTs evaluated changes in physical functioning, with mixed results: 2 studies demonstrated small improvements in the physical health component of the 36-Item Short-Form Health Survey (SF-36) and the 12-Item Short-Form Health Survey, and 2 studies demonstrated no significant difference in SF-36 scores. No studies reported on clinical outcomes related to preventive care or chronic medical care. CONCLUSIONS Integrated care models have positive effects on processes of preventive and chronic disease care but have inconsistent effects on physical functioning for individuals with serious mental illness. The relatively small number of trials and limited range of treatment models tested and outcomes reported point to the need for additional study in this important area.

Evidence Map of Yoga for Depression, Anxiety, and Posttraumatic Stress Disorder.

BACKGROUND This study describes evidence of yogas effectiveness for depressive disorders, general anxiety disorder (GAD), panic disorder (PD), and posttraumatic stress disorder (PTSD) in adults. We also address adverse events associated with yoga. METHODS We searched multiple electronic databases for systematic reviews (SRs) published between 2008 and July 2014, randomized controlled trials (RCTs) not identified in eligible SRs, and ongoing RCTs registered with ClincalTrials.gov. RESULTS We identified 1 SR on depression, 1 for adverse events, and 3 addressing multiple conditions. The high-quality depression SR included 12 RCTs (n = 619) that showed improved short-term depressive symptoms (standardized mean difference, -0.69, 95% confidence interval, -0.99 to -0.39), but there was substantial variability (I2 = 86%) and a high risk of bias for 9 studies. Three SRs addressing multiple conditions identified 4 nonrandomized studies (n = 174) for GAD/PD and 1 RCT (n = 8) and 2 nonrandomized studies (n = 22) for PTSD. We separately found 1 RCT (n = 13) for GAD and 2 RCTs (n = 102) for PTSD. Collectively, these studies were inconclusive for the effectiveness of yoga in treating GAD/PD and PTSD. The high-quality SR for adverse events included 37 primary reports (n = 76) in which inversion postures were most often implicated. We found 5 ongoing trials (3 for PTSD). CONCLUSIONS Yoga may improve short-term depressive symptoms, but evidence for GAD, PD, and PTSD remain inconclusive.

Effectiveness of Intermittent Pneumatic Compression Devices for Venous Thromboembolism Prophylaxis in High-Risk Surgical Patients: A Systematic Review

BACKGROUND Thromboprophylaxis regimens include pharmacologic and mechanical options such as intermittent pneumatic compression devices (IPCDs). There are a wide variety of IPCDs available, but it is uncertain if they vary in effectiveness or ease of use. This is a systematic review of the comparative effectiveness of IPCDs for selected outcomes (mortality, venous thromboembolism [VTE], symptomatic or asymptomatic deep vein thrombosis, major bleeding, ease of use, and adherence) in postoperative surgical patients. METHODS We searched MEDLINE (via PubMed), Embase, CINAHL, and Cochrane CENTRAL from January 1, 1995, to October 30, 2014, for randomized controlled trials, as well as relevant observational studies on ease of use and adherence. RESULTS We identified 14 eligible randomized controlled trials (2633 subjects) and 3 eligible observational studies (1724 subjects); most were conducted in joint arthroplasty patients. Intermittent pneumatic compression devices were comparable to anticoagulation for major clinical outcomes (VTE: risk ratio, 1.39; 95% confidence interval, 0.73-2.64). Limited data suggest that concurrent use of anticoagulation with IPCD may lower VTE risk compared with anticoagulation alone, and that IPCD compared with anticoagulation may lower major bleeding risk. Subgroup analyses did not show significant differences by device location, mode of inflation, or risk of bias elements. There were no consistent associations between IPCDs and ease of use or adherence. CONCLUSIONS Intermittent pneumatic compression devices are appropriate for VTE thromboprophylaxis when used in accordance with current clinical guidelines. The current evidence base to guide selection of a specific device or type of device is limited.

An evidence map of yoga for low back pain

OBJECTIVE Yoga is being increasingly studied as a treatment strategy for a variety of different clinical conditions, including low back pain (LBP). We set out to conduct an evidence map of yoga for the treatment, prevention and recurrence of acute or chronic low back pain (cLBP). METHODS We searched Medline, Cochrane Database of Systematic Reviews, EMBASE, Allied and Complementary Medicine Database and ClinicalTrials.gov for randomized controlled trials (RCT), systematic reviews or planned studies on the treatment or prevention of acute back pain or cLBP. Two independent reviewers screened papers for inclusion, extracted data and assessed the quality of included studies. RESULTS Three eligible systematic reviews were identified that included 10 RCTs (n=956) that evaluated yoga for non-specific cLBP. We did not identify additional RCTs beyond those included in the systematic reviews. Our search of ClinicalTrials.gov identified one small (n=10) unpublished trial and one large (n=320) planned clinical trial. The most recent good quality systematic review indicated significant effects for short- and long-term pain reduction (n=6 trials; standardized mean difference [SMD] -0.48; 95% CI, -0.65 to -0.31; I(2)=0% and n=5; SMD -0.33; 95% CI, -0.59 to -0.07; I(2)=48%, respectively). Long-term effects for back specific disability were also identified (n=5; SMD -0.35; 95% CI, -0.55 to -0.15; I(2)=20%). No studies were identified evaluating yoga for prevention or treatment of acute LBP. CONCLUSION Evidence suggests benefit of yoga in midlife adults with non-specific cLBP for short- and long-term pain and back-specific disability, but the effects of yoga for health-related quality of life, well- being and acute LBP are uncertain. Without additional studies, further systematic reviews are unlikely to be informative.

Nonpharmacologic, nonherbal management of menopause-associated vasomotor symptoms: an umbrella systematic review (protocol)

BackgroundVasomotor symptoms such as hot flashes and night sweats are a common concern of perimenopausal and postmenopausal women and are associated with a decreased quality of life. These symptoms can be effectively managed with hormone therapy, but safety concerns limit its use. Thus, understanding the effectiveness of nonpharmacologic therapies such as acupuncture or yoga is critical to managing these common symptoms in older women. Our review seeks to address the following question: In women with menopause-associated vasomotor symptoms, what are the effects on health-related quality of life, vasomotor symptoms, and adverse events of the following nonpharmacologic, nonherbal interventions as compared with any inactive control or active comparator: (a) acupuncture, (b) yoga, tai chi, and qigong, (c) structured exercise, and (d) meditation, mindfulness-based practices, and relaxation?MethodsWe describe a protocol for an umbrella review approach, supplemented by evaluating randomized controlled trials (RCTs) published after the most recent good-quality systematic review for each of the eligible interventions. Specific interventions were chosen based on current literature and with input from a technical expert panel and organizational stakeholders. We will conduct a thorough literature search and perform a quality assessment of potentially included systematic reviews and RCTs.DiscussionOur umbrella review, supplemented by an additional search for eligible RCTs, aims to synthesize existing evidence on the use of nonpharmacologic, nonherbal interventions to manage bothersome vasomotor symptoms in perimenopausal and postmenopausal women.Systematic review registrationPROSPERO CRD42016029335

Interventions That Support or Involve Caregivers or Families of Patients with Traumatic Injury: a Systematic Review

BackgroundAlmost 40 million family caregivers care for a loved one with severe physical or cognitive impairments. The purpose of this review is to summarize evidence about the benefits of interventions to support or involve family members/caregivers of patients with trauma-related injury on caregiver, patient, and household outcomes.MethodsEnglish-language peer-reviewed publications in MEDLINE, CINAHL, and PsycINFO from 1995 through December 2016 were identified. Eligible studies included RCT or quasi-experimental studies evaluating interventions designed to support or involve caregivers or family members of patients with TBI, PTSD, or polytrauma. Abstractions were completed by one reviewer and checked by a second; two reviewers independently assessed risk of bias using the Cochrane Effective Practice and Organization of Care Review Criteria.ResultsThirteen studies (n = 9 TBI; n = 4 PTSD, n = 0 polytrauma) evaluated psychological or rehabilitation interventions involving caregivers. Interventions did not improve TBI patients’ functional status (standardized mean difference [SMD], 0.29 [95% confidence interval [CI], − 0.51 to 1.08]) or psychological symptoms (SMD − 0.25, CI − 0.62 to 0.12). Qualitative analysis shows potential intervention benefit for TBI symptoms. Interventions did not improve TBI caregiver psychological symptoms (SMD − 0.26, CI − 0.57 to 0.05); however, qualitative analysis suggests mixed effects for caregiver burden and quality of life. Positive intervention effects on patients’ PTSD symptoms, mental health service use, and PTSD caregivers’ psychological symptoms were identified with certain interventions. Strength of evidence ranged from moderate to very low.DiscussionStudies showed mixed patterns of intervention effects on caregiver and patient outcomes; evidence about intervention impact is inconclusive. This review is the first to identify caregiving interventions for patients with TBI and polytrauma and extends past reviews about patients with PTSD. Limitations include a small evidence base, low study quality, disparate methods, varied outcome measures, and high heterogeneity. PROSPERO Registration CRD42017053516.