LETTERS
Annals of Internal Medicine COMMENTS AND RESPONSES
Masashi Miyashita, PhD Tokyo Gakugei University Tokyo, Japan
Effects of Low-Carbohydrate and Low-Fat Diets
Disclosures: Authors have disclosed no conflicts of interest. Forms
TO THE EDITOR: Bazzano and colleagues (1) confirm the find-
ings of many recent studies that show that low-carbohydrate diets are at least as effective as low-fat diets for weight loss and can reduce some cardiovascular risk factors to a greater extent. However, an integral, long-standing dietary recommendation from the American Heart Association pertaining to weight loss is to incorporate a regular pattern of physical activity (2). It is well-recognized that, despite consuming highcarbohydrate diets, endurance athletes have low levels of fasting and postprandial triglycerides, high levels of high-density lipoprotein cholesterol, and a lean body mass—factors that low-fat diets improve minimally or not at all (1). From a public health standpoint, the findings of an early randomized, controlled trial in moderately overweight and obese men and women who followed a hypocaloric National Cholesterol Education Program Step I diet of 55% carbohydrates and 30% fat for 1 year are more compelling (3). Men who walked or jogged an average of 2.1 km daily significantly decreased their triglyceride levels by 0.36 mmol/L (31.85 mg/ dL) and increased their high-density lipoprotein cholesterol levels by 0.12 mmol/L (4.63 mg/dL) more than those randomly assigned to diet only. Moreover, they lost 3.5 kg more in fat mass and had a larger reduction in systolic blood pressure. The extent of these differences was not as great in women who walked or jogged an average of 1.8 km daily; however, their high-density lipoprotein cholesterol levels remained the same over 1 year, whereas those of women who followed the diet decreased by 0.15 mmol/L (5.79 mg/dL). Subsequent work from the same group in middle-aged persons with high cholesterol levels showed that only a National Cholesterol Education Program Step II diet combined with walking or jogging approximately 16 km weekly over 1 year significantly reduced total and low-density lipoprotein cholesterol levels and that this reduction was more than that associated with diet or exercise alone (4). These interventions are supported by experimental data showing that adding 60 minutes of daily brisk walking to a short-term, low-fat diet prevented all increases in fasting and postprandial triglyceride and triglyceride-rich lipoprotein levels associated with a lowcarbohydrate diet alone in postmenopausal women (5). Collectively, such evidence shows the importance of physical activity in the American Heart Association recommendations. Bazzano and colleagues correctly suggest restricting carbohydrates as an option for weight loss, which helps reduce cardiovascular risk factors. We propose that following current American Heart Association recommendations of reducing energy and fat intake and increasing physical activity widens the choices for those who wish to lose weight but is as effective as low-carbohydrate diets for reducing weight and mitigating cardiovascular risk factors. Stephen F. Burns, PhD Nanyang Technological University Singapore
can be viewed at www.acponline.org/authors/icmje/ConflictOf InterestForms.do?msNum=L14-0508.
References 1. Bazzano LA, Hu T, Reynolds K, Yao L, Bunol C, Liu Y, et al. Effects of lowcarbohydrate and low-fat diets: a randomized trial. Ann Intern Med. 2014;161: 309-18. [PMID: 25178568] doi:10.7326/M14-0180 2. Krauss RM, Eckel RH, Howard B, Appel LJ, Daniels SR, Deckelbaum RJ, et al. AHA Dietary Guidelines: revision 2000: a statement for healthcare professionals from the Nutrition Committee of the American Heart Association. Circulation. 2000;102:2284-99. [PMID: 11056107] 3. Wood PD, Stefanick ML, Williams PT, Haskell WL. The effects on plasma lipoproteins of a prudent weight-reducing diet, with or without exercise, in overweight men and women. N Engl J Med. 1991;325:461-6. [PMID: 1852180] 4. Stefanick ML, Mackey S, Sheehan M, Ellsworth N, Haskell WL, Wood PD. Effects of diet and exercise in men and postmenopausal women with low levels of HDL cholesterol and high levels of LDL cholesterol. N Engl J Med. 1998; 339:12-20. [PMID: 9647874] 5. Koutsari C, Karpe F, Humphreys SM, Frayn KN, Hardman AE. Exercise prevents the accumulation of triglyceride-rich lipoproteins and their remnants seen when changing to a high-carbohydrate diet. Arterioscler Thromb Vasc Biol. 2001;21:1520-5. [PMID: 11557682]
TO THE EDITOR: In Bazzano and colleagues' trial of 148 obese
patients carried out in a single medical center (1), a lowcarbohydrate ( 0.001) (1). The original Framingham Study revealed that the risk for coronary artery disease increases sharply as high-density lipoprotein cholesterol levels decrease progressively less than 40.00 mg/dL (1.03 mmol/L), which supports this finding (3). Equally important, approximately one third of the population enrolled in Bazzano and colleagues' trial was receiving antihypertensive treatment that was not specified (1). Given the complex relationship between endothelial dysfunction and high-density lipoprotein cholesterol levels (4, 5), it would be noteworthy to know the results of the analysis distinguishing between hypertensive and nonhypertensive participants to explore new pathophysiologic insights on the association linking hypertension and obesity.
Gaetano Santulli, MD, PhD Columbia University Medical Center College of Physicians and Surgeons New York, New York
4. Bønaa KH, Thelle DS. Association between blood pressure and serum lipids in a population. The Tromsø Study. Circulation. 1991;83:1305-14. [PMID: 2013148] 5. Yuhanna IS, Zhu Y, Cox BE, Hahner LD, Osborne-Lawrence S, Lu P, et al. High-density lipoprotein binding to scavenger receptor-BI activates endothelial nitric oxide synthase. Nat Med. 2001;7:853-7. [PMID: 11433352]
TO THE EDITOR: Bazzano and colleagues (1) hypothesized that participants following a low-carbohydrate diet would have lower body weight than those adhering to a low-fat diet and conclude that the results of their trial confirm this hypothesis. Neither diet included a specific calorie or energy intake goal; however, it seems that the behavioral curriculum followed by participants targeted weight loss because reported energy intake decreased from baseline during the intervention in the low-carbohydrate and low-fat diet groups. Mean weight loss was greater among participants assigned to a lowcarbohydrate diet. However, we estimate that underreporting of dietary energy intake was close to 50% in both groups during the intervention. Mean daily energy intake at 3-, 6-, and 12-month follow-up were reported to be between 1258 and 1527 kcal. A full year of such a low energy intake should have led to weight loss in excess of 30 kg rather than the 2 to 5 kg that took place. The degree of underreporting evident from these results makes it very difficult to know what participants actually ate and, more important, creates considerable uncertainty about the extent to which the greater weight loss in one diet group can be attributed to the macronutrient composition of the diet. This severe disconnection between reported and actual energy intake begs the question about the composition of the unreported and underreported foods and suggests that forces other than the macronutrient composition of the diet may explain the weight differences. As such, the lack of blinding in weight loss trials makes it particularly difficult to disentangle the effect of macronutrients from the behaviors that the diets may invoke. Katan remarked in an editorial several years ago (2) that participants in weight loss trials “may eat less not because of the protein or carbohydrate content of a diet but because of the diet's reputation or novelty or because of the taste of particular foods in the diet.” The widespread promotion of low-carbohydrate diets may well explain these findings. A series of meta-analyses, including one published recently (3), showed little to differentiate the long-term effects on body weight of various dietary patterns intended to
392 Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
www.annals.org
LETTERS achieve weight loss. Bazzano and colleagues' study certainly does not refute this conclusion. Murray Skeaff, BSc(Hons), PhD Jim Mann, MBChB, PhD, MA, DM Lisa Te Morenga, BForSc, BSc, PhD Rachael McLean, BA, MBChB, PhD University of Otago Dunedin, New Zealand Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L14-0506. References 1. Bazzano LA, Hu T, Reynolds K, Yao L, Bunol C, Liu Y, et al. Effects of lowcarbohydrate and low-fat diets: a randomized trial. Ann Intern Med. 2014;161: 309-18. [PMID: 25178568] doi:10.7326/M14-0180 2. Katan MB. Weight-loss diets for the prevention and treatment of obesity [Editorial]. N Engl J Med. 2009;360:923-5. [PMID: 19246365] doi:10.1056 /NEJMe0810291 3. Naude CE, Schoonees A, Senekal M, Young T, Garner P, Volmink J. Low carbohydrate versus isoenergetic balanced diets for reducing weight and cardiovascular risk: a systematic review and meta-analysis. PLoS One. 2014;9: e100652. [PMID: 25007189] doi:10.1371/journal.pone.0100652
IN RESPONSE: We agree wholeheartedly with Drs. Burns and
Miyashita that physical activity is a vital component of weight management and cardiovascular health. In our clinical trial, physical activity levels did not significantly differ between the groups at baseline or during the intervention. Although it was not the aim of our study, future studies should examine whether physical activity levels modify the differences in efficacy of low-fat and low-carbohydrate diets. Drs. Donzelli and Lafranconi raise several concerns about low-carbohydrate diets. They note that a meta-analysis showed that low-carbohydrate and low-fat diets were similarly effective for weight loss at 12-month follow-up (1). However, unlike our study, Johnston and colleagues' review—which focused on such named low-carbohydrate diets as Atkins and Zone— did not examine cardiovascular disease risk factors. They also point out that a Mediterranean diet may have advantages over a low-carbohydrate diet (2). Although our study was not designed to examine this question, in DIRECT, Mediterranean and low-carbohydrate diet groups experienced more favorable postintervention effects than low-fat diet group. Drs. Donzelli and Lafranconi note that lowcarbohydrate diets may be associated with higher adult body mass index, all-cause mortality, or both in observational studies. However, evidence from clinical trials does not support an increased body mass index (1–3), and many observational cohort studies have identified no association between lowcarbohydrate diets and mortality (4, 5). Finally, Drs. Donzelli and Lafranconi state that low-carbohydrate and high-protein diets are environmentally unsustainable at a population level. We disagree. Proteins from legumes, nuts, and seeds provide a potentially sustainable pathway to diets higher in protein and healthy fats. Dr. Santulli stated that weight loss differed significantly only at 3- and 6-month follow-up but not at 12 months and that the only variables that differed significantly between the 2 diets at 12-month follow-up were high-density lipoprotein www.annals.org
cholesterol levels and 10-year Framingham Risk Score; we disagree. In fact, Table 3 in our article shows that changes in body weight, lean mass, fat mass, total– high-density lipoprotein cholesterol ratio, and C-reactive protein levels also differed significantly between the groups at 12 months. The Framingham Risk Score differed significantly at 3, 6, and 12 months. Sixty of our participants were hypertensive. The numbers of hypertensive and normotensive participants did not differ significantly by diet group. Dr. Skeaff and colleagues' concern about underreporting in our trial was based on mean daily energy intake reported at 3, 6, and 12 months. These quantities concur with the caloric intake and weight loss results reported by 23 studies of lowfat and low-carbohydrate diets in a meta-analysis of randomized, controlled trials (3). Participants in behavioral dietary trials are free-living volunteers from the community, and their exact caloric intake and output cannot be measured in the idealized manner of physiologic studies done on metabolic wards. In terms of the speculation that the widespread promotion of low-carbohydrate diets for weight loss may have influenced a difference in weight loss between the diets, we argue that low-fat diets have been far more heavily promoted for many decades than low-carbohydrate diets. Lydia A. Bazzano, MD, PhD, MPH Tian Hu, MD, MS Tulane University School of Public Health and Tropical Medicine New Orleans, Louisiana Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=M14-0180. References 1. Johnston BC, Kanters S, Bandayrel K, Wu P, Naji F, Siemieniuk RA, et al. Comparison of weight loss among named diet programs in overweight and obese adults: a meta-analysis. JAMA. 2014;312:923-33. [PMID: 25182101] doi:10.1001/jama.2014.10397 2. Schwarzfuchs D, Golan R, Shai I. Four-year follow-up after two-year dietary interventions [Letter]. N Engl J Med. 2012;367:1373-4. [PMID: 23034044] doi: 10.1056/NEJMc1204792 3. Hu T, Mills KT, Yao L, Demanelis K, Eloustaz M, Yancy WS Jr, et al. Effects of low-carbohydrate diets versus low-fat diets on metabolic risk factors: a metaanalysis of randomized controlled clinical trials. Am J Epidemiol. 2012;176 Suppl 7:S44-54. [PMID: 23035144] doi:10.1093/aje/kws264 4. Fung TT, van Dam RM, Hankinson SE, Stampfer M, Willett WC, Hu FB. Lowcarbohydrate diets and all-cause and cause-specific mortality: two cohort studies. Ann Intern Med. 2010;153:289-98. [PMID: 20820038] doi:10.7326/0003 -4819-153-5-201009070-00003 5. Nilsson LM, Winkvist A, Eliasson M, Jansson JH, Hallmans G, Johansson I, et al. Low-carbohydrate, high-protein score and mortality in a northern Swedish population-based cohort. Eur J Clin Nutr. 2012;66:694-700. [PMID: 22333874] doi:10.1038/ejcn.2012.9
Treatment of Latent Tuberculosis Infection TO THE EDITOR: In their network meta-analysis assessing regimens for the treatment of latent tuberculosis infection, Stagg and colleagues (1) concluded that therapies containing rifamycins may be effective alternatives to isoniazid monotherapy. Many aspects of their study are worth commendaAnnals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015 393
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
LETTERS tion, including a broad search strategy, collection and adjustment of covariates for the development of active tuberculosis, and comprehensive reporting of outcomes in their article and the supplementary material. The authors discussed the inconsistency of the higher rate of hepatotoxicity in their placebo group and attributed this occurrence to the rate of hepatic events in the placebo groups of some of the individual trials. Of note, the risk for active tuberculosis differed significantly between the placebo and no-treatment groups. This finding suggests potential heterogeneity in the diagnostic methods and incidence of tuberculosis between trials that used placebo or no treatment as their control. In addition, their final analysis showed that treatment with isoniazid for 9 months, a currently recommended regimen for latent tuberculosis infection, did not decrease the risk for active tuberculosis (odds ratio, 0.94 [95% credible interval, 0.40 to 2.10]) compared with placebo. When faced with such inconsistencies, questions about the credibility of the network should be raised. Might the inclusion of poorer-quality trials have adversely affected the network? Many poor-quality trials escaped exclusion criteria (defined as high risk of bias across 4 categories) because bias was not able to be classified across 1 or more categories and was therefore deemed unclear. To their credit, the authors reported a subsequent analysis showing the absence of significant effect modification of trials with unclear or high risk of bias. However, because such a study may inform global policy, perhaps more stringent exclusion criteria should apply. These criteria may not only eliminate the inconsistencies listed here but also further reassure the reader about the equivalency between rifamycin-based regimens and isoniazid monotherapy. Reuben Arasaratnam, MD, MPH Baylor College of Medicine Houston, Texas Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOf InterestForms.do?msNum=L14-0554. Reference 1. Stagg HR, Zenner D, Harris RJ, Mun˜oz L, Lipman MC, Abubakar I. Treatment of latent tuberculosis infection: a network meta-analysis. Ann Intern Med. 2014;161:419-28. [PMID: 25111745] doi:10.7326/M14-1019
TO THE EDITOR: I read Stagg and colleagues' (1) review with
great interest. As far as I know, production of isolated isoniazid has been discontinued. However, the high prevalence of latent tuberculosis infection in many countries has necessitated the return of this preparation to the market because it is the only regimen for the management of this condition that is highly efficacious as well as cost-effective. Gauranga C. Dhar, MD Bangladesh Institute of Family Medicine and Research Dhaka, Bangladesh Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOf InterestForms.do?msNum=L14-0555.
Reference 1. Stagg HR, Zenner D, Harris RJ, Mun˜oz L, Lipman MC, Abubakar I. Treatment of latent tuberculosis infection: a network meta-analysis. Ann Intern Med. 2014;161:419-28. [PMID: 25111745] doi:10.7326/M14-1019
IN RESPONSE: Dr. Arasaratnam raised several issues about
our systematic review, which was limited by the availability and quality of underlying articles because the literature for most latent tuberculosis infection regimens is relatively sparse. After rigorously assessing study quality, we tested for associations between quality indicators and outcomes but did not find particularly strong evidence for such associations; thus, this influence was likely minimal. Given the small number and frequently poor or unassessable quality of trials, further exclusion of studies would have limited, rather than provided assurance about, the comparison between rifamycins and isoniazid. Network meta-analyses pose a risk that indirect findings could be inconsistent with direct pairwise comparisons. We examined this possibility but did not find evidence to invalidate the network approach, although low statistical power was a limitation. To further validate our results, our parallel conventional meta-analysis showed similar effect estimates compared with the network analysis. Study quality would be of particular concern if it led to exaggeration of summary estimates (for example, an increase in efficacy). Varying study quality is likely to increase heterogeneity rather than induce bias in a particular direction. Substantial heterogeneity was observed in our study, leading to wider credible intervals and difficulties determining true differences in treatment effect. To some extent, we accounted for the heterogeneity by using a random-effects model. Despite the limitations of the included studies, evidence is available on the efficacy and safety of rifampicin regimens, including 3 to 4 months of monotherapy (1). Another recent meta-analysis also provides support for rifamycin-based regimens (2). Our systematic review supports the evidence that isoniazid for 6 months or longer is effective and safe. The observed limited efficacy of isoniazid monotherapy for 9 months is likely due to paucity of data. Previous analysis of trial data by Comstock suggests that the greatest efficacy of isoniazid regimens is likely achieved after 6 to 12 months of treatment (3). We agree with Dr. Dhar that isoniazid monotherapy remains an important treatment option to prevent progression from latent tuberculosis infection to active tuberculosis (4). The availability of appropriate isoniazid formulations therefore needs to be globally assured. The limitations outlined in our review indicate that further studies are needed to reliably evaluate the efficacy of treatment regimens for latent tuberculosis infection. Dominik Zenner, MD University College London and Public Health England London, United Kingdom Helen R. Stagg, PhD University College London London, United Kingdom
394 Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
www.annals.org
LETTERS Ross J. Harris, MSc Public Health England London, United Kingdom Ibrahim Abubakar, MBBS, PhD University College London and Public Health England London, United Kingdom Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=M14-1019.
Table. Correlation Between Confidence and Accuracy in Men and Women Confidence Rating
“I'm sure” “Feeling lucky” “No clue”
Confidence for All Questions Answered, %
Accuracy by Confidence Level, %
Men
Women
P Value*
Men
Women
P Value*
44.4 42.0 13.6
39.5 44.4 16.1
>0.001 0.180 0.27
78.3 49.8 32.0
80.5 53.5 32.0
>0.002 >0.001 1
* From Mann–Whitney U tests with Bonferroni adjustment.
References 1. A double-blind placebo-controlled clinical trial of three antituberculosis chemoprophylaxis regimens in patients with silicosis in Hong Kong. Hong Kong Chest Service/Tuberculosis Research Centre, Madras/British Medical Research Council. Am Rev Respir Dis. 1992;145:36-41. [PMID: 1731596] 2. Sharma SK, Sharma A, Kadhiravan T, Tharyan P. Rifamycins (rifampicin, rifabutin and rifapentine) compared to isoniazid for preventing tuberculosis in HIV-negative people at risk of active TB. Cochrane Database Syst Rev. 2013;7:CD007545. [PMID: 23828580] doi:10.1002/14651858.CD007545 .pub2 3. Comstock GW. How much isoniazid is needed for prevention of tuberculosis among immunocompetent adults? Int J Tuberc Lung Dis. 1999;3:847-50. [PMID: 10524579] 4. World Health Organization. Guidelines on the Management of Latent Tuberculosis Infection. Geneva: World Health Organization; 2014. Accessed at www.who.int/tb/publications/ltbi_document_page/en/ on 24 January 2015.
OBSERVATIONS The Association Between Confidence and Accuracy Among Users of a Mobile Web Platform for Medical Education Background: Recent literature suggests that physicians' diagnostic confidence tends to exceed accuracy, with confidence being relatively inflexible regardless of case difficulty (1). Overconfidence could lead to physicians not asking for help when they need it. Underconfidence may expose patients to defensive medicine. Those who lack confidence may misrepresent their actual level of knowledge, resulting in missed opportunities for professional development (2). Research suggests that female medical students perform as well as their male peers but report less confidence in their abilities and are actually perceived to be less confident (3). The future of medical education may involve training students to better align confidence and accuracy. One strategy may include providing students with immediate, actionable performance feedback through technology-enabled education platforms (4). Osmosis is a mobile Web platform designed to help medical students learn by answering questions related to their curricula. Its database includes information from more than 14 000 users. We developed a user-facing “Calibration Index” to show students their propensity for under- or overconfidence when answering questions. Objective: To measure medical students' confidence in their answers, compare confidence with accuracy, and examine differences between men and women. Methods: To answer a question in the platform, a user must first select 1 of 3 confidence ratings: “I'm sure,” “Feeling lucky,” or “No clue.” We analyzed the responses of 1021 users who answered more than 50 random multiple-choice and www.annals.org
true-or-false questions. An automated software package (SexMachine, version 0.1.1, based on Gender.c, version 1.2) that evaluates data on how frequently a name correlates with male or female gender in the United States was used to determine a user's sex from his or her first name. Users with ambiguous name–sex pairings were omitted. The final list was checked to confirm reasonable assignment. Users were assumed to be predominantly medical students in their preclinical years because the platform is designed for and marketed to this audience. Results are presented as the means of user-level response averages and compared with Mann–Whitney U tests using the Bonferroni adjustment (Stata, version 12.0, StataCorp). Findings: The study included 617 men and 404 women. Women's answers were more accurate than men's (61.4% vs. 60.3% correct; P = 0.040); however, they were less confident than men. For example, women selected “I'm sure” significantly less often than men (39.5% vs. 44.4% of responses) (Table). Average accuracy for questions rated as “I'm sure” and “Feeling lucky” was higher in women (80.5% vs. 78.3% of “I'm sure” responses were correct; 53.5% vs. 49.8% of “Feeling lucky” responses were correct). The average number of questions answered by female and male users did not differ (647 vs. 579; P = 0.40). When results were stratified into quartiles based on the number of questions answered, differences by sex remained. Discussion: The data suggest that, despite performing at the same level or higher, women lacked confidence relative to men. To understand the meaning and magnitude of this observed effect, it will be important to relate question-answering behavior to actual behavior on the wards. One limitation is that sex was determined from first name. Also, we could not verify that users were medical students or stratify results according to school and academic year. Notwithstanding these limitations, the data are a reminder that less confidence might not indicate a lack of knowledge and that confidence should not be mistaken for correctness. Insights gained from understanding the relationship between confidence and accuracy in medical trainees may ultimately prove valuable in reducing diagnostic errors caused by overconfidence and sex disparities caused by lack of confidence. Going forward, new educational interventions like the Calibration Index may help future physicians align confidence with accuracy to improve patient care and promote career advancement (5). Jason Theobald, BA New York University School of Medicine New York, New York Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015 395
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
LETTERS Shiv Gaglani, BA M. Ryan Haynes, BS, MPhil, PhD Johns Hopkins School of Medicine Baltimore, Maryland Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L14-0470. References 1. Meyer AN, Payne VL, Meeks DW, Rao R, Singh H. Physicians' diagnostic accuracy, confidence, and resource requests: a vignette study. JAMA Intern Med. 2013;173:1952-8. [PMID: 23979070] doi:10.1001/jamainternmed.2013 .10081 2. Nomura K, Yano E, Fukui T. Gender differences in clinical confidence: a nationwide survey of resident physicians in Japan. Acad Med. 2010;85:647-53. [PMID: 20354381] doi:10.1097/ACM.0b013e3181d2a796 3. Blanch DC, Hall JA, Roter DL, Frankel RM. Medical student gender and issues of confidence. Patient Educ Couns. 2008;72:374-81. [PMID: 18656322] doi:10.1016/j.pec.2008.05.021 4. Berner ES, Graber ML. Overconfidence as a cause of diagnostic error in medicine. Am J Med. 2008;121:S2-23. [PMID: 18440350] doi:10.1016 /j.amjmed.2008.01.001 5. Gaglani SM, Haynes MR. What can medical education learn from Facebook and Netflix? Ann Intern Med. 2014;160:640-1. [PMID: 24798524] doi:10.7326 /M13-2286
Progress and Challenges in Electronic Health Record Adoption: Findings From a National Survey of Physicians Background: The United States is modernizing the information technology infrastructure of its health care system. The provisions of the American Recovery and Reinvestment Act that are known as the Health Information Technology for Economic and Clinical Health (HITECH) Act provide substantial economic incentives for physicians to implement and use electronic health records (EHRs) (1, 2). Since the passage of the HITECH Act, studies have consistently found EHR use increasing steadily among office-based physicians (3, 4). However, recent studies report that a substantial proportion of physicians are unsure about or are not planning to participate in the program (5). Understanding which physicians may opt out will be necessary to achieve the goals of the HITECH Act and implement broader changes. Methods: We surveyed a sample of 3437 U.S. physicians selected to represent all primary care physicians and certain specialist physicians who were likely to care for a given patient over an extended time. We collected data in 2 waves. Wave 1 took place between October 2011 and March 2012, and wave 2 took place between May and July 2013. Detailed information about the survey methods is available from the authors on request. In this observation, we describe data from physicians who responded to both rounds of the survey (44% overall response rate). Our primary goal was to measure associations between the stage of EHR adoption and practice characteristics. Findings: In 2011, 44% of physicians had an EHR that met basic criteria (early adopters). Between 2011 and 2013, an additional 19% adopted a basic EHR (new adopters). In 2013, 20% were in the process of implementing an EHR or had implemented one without some functions required for a basic
EHR (partial implementers), 8% were planning to adopt an EHR in the next 2 years (planners), and 9% were not planning to adopt an EHR (persistent nonadopters). On average, persistent nonadopters were older than other physicians. The mean number of physicians employed in the main practice location of persistent nonadopters was 2.3 compared with 33.4 among early adopters and 15.1 among new adopters. Persistent nonadopters were significantly more likely to be employed in independent solo or 2-physician practices than early and new adopters, who were significantly more likely to be employed by a hospital or medical school, group or staff-model HMO, or network owned by a hospital or other type of health care organization. Methods of compensation varied by stage of implementation. Most persistent nonadopters reported fee-for-service as their primary compensation, whereas early and new adopters were more likely to report salary adjusted for performance. Persistent nonadopters seemed less likely to participate in incentive programs focused on improving the quality and continuity of care and were significantly less likely than early and new adopters to receive or have the potential to receive additional payments for managing patients with chronic conditions or complex needs. Discussion: Persistent nonadopters in small, isolated practices may be facing a unique set of challenges that limits their ability to adopt an EHR. Failure to address the needs of these physicians has implications beyond adoption because new models of health care delivery require the use of an EHR. Physicians who choose not to make the change to EHRs may find themselves further isolated if these new models become widespread, but they may move toward adoption as the penalty phase of the meaningful use program draws closer. If so, they are likely to require extensive support in selecting, implementing, and using these systems. Catherine DesRoches, PhD Mathematica Policy Research Cambridge, Massachusetts Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L14-0466. References 1. The American Recovery and Reinvestment Act, Pub L. No. 111-5, Stat. 123 (2009). 2. Blumenthal D. Wiring the health system— origins and provisions of a new federal program. N Engl J Med. 2011;365:2323-9. [PMID: 22168647] doi: 10.1056/NEJMsr1110507 3. Furukawa MF, King J, Patel V, Hsiao CJ, Adler-Milstein J, Jha AK. Despite substantial progress in EHR adoption, health information exchange and patient engagement remain low in office settings. Health Aff (Millwood). 2014; 33:1672-9. [PMID: 25104827] doi:10.1377/hlthaff.2014.0445 4. Audet AM, Squires D, Doty MM. Where are we on the diffusion curve? Trends and drivers of primary care physicians' use of health information technology. Health Serv Res. 2014;49:347-60. [PMID: 24358958] doi:10.1111 /1475-6773.12139 5. Hsiao C, Hing E. Use and Characteristics of Electronic Health Record Systems Among Office-based Physician Practices: United States, 2001-2013. NCHS Data Brief no. 143. Hyattsville, MD: National Center for Health Statistics; 2014.
396 Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
www.annals.org
Annals of Internal Medicine COMMENTS AND RESPONSES
Masashi Miyashita, PhD Tokyo Gakugei University Tokyo, Japan
Effects of Low-Carbohydrate and Low-Fat Diets
Disclosures: Authors have disclosed no conflicts of interest. Forms
TO THE EDITOR: Bazzano and colleagues (1) confirm the find-
ings of many recent studies that show that low-carbohydrate diets are at least as effective as low-fat diets for weight loss and can reduce some cardiovascular risk factors to a greater extent. However, an integral, long-standing dietary recommendation from the American Heart Association pertaining to weight loss is to incorporate a regular pattern of physical activity (2). It is well-recognized that, despite consuming highcarbohydrate diets, endurance athletes have low levels of fasting and postprandial triglycerides, high levels of high-density lipoprotein cholesterol, and a lean body mass—factors that low-fat diets improve minimally or not at all (1). From a public health standpoint, the findings of an early randomized, controlled trial in moderately overweight and obese men and women who followed a hypocaloric National Cholesterol Education Program Step I diet of 55% carbohydrates and 30% fat for 1 year are more compelling (3). Men who walked or jogged an average of 2.1 km daily significantly decreased their triglyceride levels by 0.36 mmol/L (31.85 mg/ dL) and increased their high-density lipoprotein cholesterol levels by 0.12 mmol/L (4.63 mg/dL) more than those randomly assigned to diet only. Moreover, they lost 3.5 kg more in fat mass and had a larger reduction in systolic blood pressure. The extent of these differences was not as great in women who walked or jogged an average of 1.8 km daily; however, their high-density lipoprotein cholesterol levels remained the same over 1 year, whereas those of women who followed the diet decreased by 0.15 mmol/L (5.79 mg/dL). Subsequent work from the same group in middle-aged persons with high cholesterol levels showed that only a National Cholesterol Education Program Step II diet combined with walking or jogging approximately 16 km weekly over 1 year significantly reduced total and low-density lipoprotein cholesterol levels and that this reduction was more than that associated with diet or exercise alone (4). These interventions are supported by experimental data showing that adding 60 minutes of daily brisk walking to a short-term, low-fat diet prevented all increases in fasting and postprandial triglyceride and triglyceride-rich lipoprotein levels associated with a lowcarbohydrate diet alone in postmenopausal women (5). Collectively, such evidence shows the importance of physical activity in the American Heart Association recommendations. Bazzano and colleagues correctly suggest restricting carbohydrates as an option for weight loss, which helps reduce cardiovascular risk factors. We propose that following current American Heart Association recommendations of reducing energy and fat intake and increasing physical activity widens the choices for those who wish to lose weight but is as effective as low-carbohydrate diets for reducing weight and mitigating cardiovascular risk factors. Stephen F. Burns, PhD Nanyang Technological University Singapore
can be viewed at www.acponline.org/authors/icmje/ConflictOf InterestForms.do?msNum=L14-0508.
References 1. Bazzano LA, Hu T, Reynolds K, Yao L, Bunol C, Liu Y, et al. Effects of lowcarbohydrate and low-fat diets: a randomized trial. Ann Intern Med. 2014;161: 309-18. [PMID: 25178568] doi:10.7326/M14-0180 2. Krauss RM, Eckel RH, Howard B, Appel LJ, Daniels SR, Deckelbaum RJ, et al. AHA Dietary Guidelines: revision 2000: a statement for healthcare professionals from the Nutrition Committee of the American Heart Association. Circulation. 2000;102:2284-99. [PMID: 11056107] 3. Wood PD, Stefanick ML, Williams PT, Haskell WL. The effects on plasma lipoproteins of a prudent weight-reducing diet, with or without exercise, in overweight men and women. N Engl J Med. 1991;325:461-6. [PMID: 1852180] 4. Stefanick ML, Mackey S, Sheehan M, Ellsworth N, Haskell WL, Wood PD. Effects of diet and exercise in men and postmenopausal women with low levels of HDL cholesterol and high levels of LDL cholesterol. N Engl J Med. 1998; 339:12-20. [PMID: 9647874] 5. Koutsari C, Karpe F, Humphreys SM, Frayn KN, Hardman AE. Exercise prevents the accumulation of triglyceride-rich lipoproteins and their remnants seen when changing to a high-carbohydrate diet. Arterioscler Thromb Vasc Biol. 2001;21:1520-5. [PMID: 11557682]
TO THE EDITOR: In Bazzano and colleagues' trial of 148 obese
patients carried out in a single medical center (1), a lowcarbohydrate ( 0.001) (1). The original Framingham Study revealed that the risk for coronary artery disease increases sharply as high-density lipoprotein cholesterol levels decrease progressively less than 40.00 mg/dL (1.03 mmol/L), which supports this finding (3). Equally important, approximately one third of the population enrolled in Bazzano and colleagues' trial was receiving antihypertensive treatment that was not specified (1). Given the complex relationship between endothelial dysfunction and high-density lipoprotein cholesterol levels (4, 5), it would be noteworthy to know the results of the analysis distinguishing between hypertensive and nonhypertensive participants to explore new pathophysiologic insights on the association linking hypertension and obesity.
Gaetano Santulli, MD, PhD Columbia University Medical Center College of Physicians and Surgeons New York, New York
4. Bønaa KH, Thelle DS. Association between blood pressure and serum lipids in a population. The Tromsø Study. Circulation. 1991;83:1305-14. [PMID: 2013148] 5. Yuhanna IS, Zhu Y, Cox BE, Hahner LD, Osborne-Lawrence S, Lu P, et al. High-density lipoprotein binding to scavenger receptor-BI activates endothelial nitric oxide synthase. Nat Med. 2001;7:853-7. [PMID: 11433352]
TO THE EDITOR: Bazzano and colleagues (1) hypothesized that participants following a low-carbohydrate diet would have lower body weight than those adhering to a low-fat diet and conclude that the results of their trial confirm this hypothesis. Neither diet included a specific calorie or energy intake goal; however, it seems that the behavioral curriculum followed by participants targeted weight loss because reported energy intake decreased from baseline during the intervention in the low-carbohydrate and low-fat diet groups. Mean weight loss was greater among participants assigned to a lowcarbohydrate diet. However, we estimate that underreporting of dietary energy intake was close to 50% in both groups during the intervention. Mean daily energy intake at 3-, 6-, and 12-month follow-up were reported to be between 1258 and 1527 kcal. A full year of such a low energy intake should have led to weight loss in excess of 30 kg rather than the 2 to 5 kg that took place. The degree of underreporting evident from these results makes it very difficult to know what participants actually ate and, more important, creates considerable uncertainty about the extent to which the greater weight loss in one diet group can be attributed to the macronutrient composition of the diet. This severe disconnection between reported and actual energy intake begs the question about the composition of the unreported and underreported foods and suggests that forces other than the macronutrient composition of the diet may explain the weight differences. As such, the lack of blinding in weight loss trials makes it particularly difficult to disentangle the effect of macronutrients from the behaviors that the diets may invoke. Katan remarked in an editorial several years ago (2) that participants in weight loss trials “may eat less not because of the protein or carbohydrate content of a diet but because of the diet's reputation or novelty or because of the taste of particular foods in the diet.” The widespread promotion of low-carbohydrate diets may well explain these findings. A series of meta-analyses, including one published recently (3), showed little to differentiate the long-term effects on body weight of various dietary patterns intended to
392 Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
www.annals.org
LETTERS achieve weight loss. Bazzano and colleagues' study certainly does not refute this conclusion. Murray Skeaff, BSc(Hons), PhD Jim Mann, MBChB, PhD, MA, DM Lisa Te Morenga, BForSc, BSc, PhD Rachael McLean, BA, MBChB, PhD University of Otago Dunedin, New Zealand Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L14-0506. References 1. Bazzano LA, Hu T, Reynolds K, Yao L, Bunol C, Liu Y, et al. Effects of lowcarbohydrate and low-fat diets: a randomized trial. Ann Intern Med. 2014;161: 309-18. [PMID: 25178568] doi:10.7326/M14-0180 2. Katan MB. Weight-loss diets for the prevention and treatment of obesity [Editorial]. N Engl J Med. 2009;360:923-5. [PMID: 19246365] doi:10.1056 /NEJMe0810291 3. Naude CE, Schoonees A, Senekal M, Young T, Garner P, Volmink J. Low carbohydrate versus isoenergetic balanced diets for reducing weight and cardiovascular risk: a systematic review and meta-analysis. PLoS One. 2014;9: e100652. [PMID: 25007189] doi:10.1371/journal.pone.0100652
IN RESPONSE: We agree wholeheartedly with Drs. Burns and
Miyashita that physical activity is a vital component of weight management and cardiovascular health. In our clinical trial, physical activity levels did not significantly differ between the groups at baseline or during the intervention. Although it was not the aim of our study, future studies should examine whether physical activity levels modify the differences in efficacy of low-fat and low-carbohydrate diets. Drs. Donzelli and Lafranconi raise several concerns about low-carbohydrate diets. They note that a meta-analysis showed that low-carbohydrate and low-fat diets were similarly effective for weight loss at 12-month follow-up (1). However, unlike our study, Johnston and colleagues' review—which focused on such named low-carbohydrate diets as Atkins and Zone— did not examine cardiovascular disease risk factors. They also point out that a Mediterranean diet may have advantages over a low-carbohydrate diet (2). Although our study was not designed to examine this question, in DIRECT, Mediterranean and low-carbohydrate diet groups experienced more favorable postintervention effects than low-fat diet group. Drs. Donzelli and Lafranconi note that lowcarbohydrate diets may be associated with higher adult body mass index, all-cause mortality, or both in observational studies. However, evidence from clinical trials does not support an increased body mass index (1–3), and many observational cohort studies have identified no association between lowcarbohydrate diets and mortality (4, 5). Finally, Drs. Donzelli and Lafranconi state that low-carbohydrate and high-protein diets are environmentally unsustainable at a population level. We disagree. Proteins from legumes, nuts, and seeds provide a potentially sustainable pathway to diets higher in protein and healthy fats. Dr. Santulli stated that weight loss differed significantly only at 3- and 6-month follow-up but not at 12 months and that the only variables that differed significantly between the 2 diets at 12-month follow-up were high-density lipoprotein www.annals.org
cholesterol levels and 10-year Framingham Risk Score; we disagree. In fact, Table 3 in our article shows that changes in body weight, lean mass, fat mass, total– high-density lipoprotein cholesterol ratio, and C-reactive protein levels also differed significantly between the groups at 12 months. The Framingham Risk Score differed significantly at 3, 6, and 12 months. Sixty of our participants were hypertensive. The numbers of hypertensive and normotensive participants did not differ significantly by diet group. Dr. Skeaff and colleagues' concern about underreporting in our trial was based on mean daily energy intake reported at 3, 6, and 12 months. These quantities concur with the caloric intake and weight loss results reported by 23 studies of lowfat and low-carbohydrate diets in a meta-analysis of randomized, controlled trials (3). Participants in behavioral dietary trials are free-living volunteers from the community, and their exact caloric intake and output cannot be measured in the idealized manner of physiologic studies done on metabolic wards. In terms of the speculation that the widespread promotion of low-carbohydrate diets for weight loss may have influenced a difference in weight loss between the diets, we argue that low-fat diets have been far more heavily promoted for many decades than low-carbohydrate diets. Lydia A. Bazzano, MD, PhD, MPH Tian Hu, MD, MS Tulane University School of Public Health and Tropical Medicine New Orleans, Louisiana Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=M14-0180. References 1. Johnston BC, Kanters S, Bandayrel K, Wu P, Naji F, Siemieniuk RA, et al. Comparison of weight loss among named diet programs in overweight and obese adults: a meta-analysis. JAMA. 2014;312:923-33. [PMID: 25182101] doi:10.1001/jama.2014.10397 2. Schwarzfuchs D, Golan R, Shai I. Four-year follow-up after two-year dietary interventions [Letter]. N Engl J Med. 2012;367:1373-4. [PMID: 23034044] doi: 10.1056/NEJMc1204792 3. Hu T, Mills KT, Yao L, Demanelis K, Eloustaz M, Yancy WS Jr, et al. Effects of low-carbohydrate diets versus low-fat diets on metabolic risk factors: a metaanalysis of randomized controlled clinical trials. Am J Epidemiol. 2012;176 Suppl 7:S44-54. [PMID: 23035144] doi:10.1093/aje/kws264 4. Fung TT, van Dam RM, Hankinson SE, Stampfer M, Willett WC, Hu FB. Lowcarbohydrate diets and all-cause and cause-specific mortality: two cohort studies. Ann Intern Med. 2010;153:289-98. [PMID: 20820038] doi:10.7326/0003 -4819-153-5-201009070-00003 5. Nilsson LM, Winkvist A, Eliasson M, Jansson JH, Hallmans G, Johansson I, et al. Low-carbohydrate, high-protein score and mortality in a northern Swedish population-based cohort. Eur J Clin Nutr. 2012;66:694-700. [PMID: 22333874] doi:10.1038/ejcn.2012.9
Treatment of Latent Tuberculosis Infection TO THE EDITOR: In their network meta-analysis assessing regimens for the treatment of latent tuberculosis infection, Stagg and colleagues (1) concluded that therapies containing rifamycins may be effective alternatives to isoniazid monotherapy. Many aspects of their study are worth commendaAnnals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015 393
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
LETTERS tion, including a broad search strategy, collection and adjustment of covariates for the development of active tuberculosis, and comprehensive reporting of outcomes in their article and the supplementary material. The authors discussed the inconsistency of the higher rate of hepatotoxicity in their placebo group and attributed this occurrence to the rate of hepatic events in the placebo groups of some of the individual trials. Of note, the risk for active tuberculosis differed significantly between the placebo and no-treatment groups. This finding suggests potential heterogeneity in the diagnostic methods and incidence of tuberculosis between trials that used placebo or no treatment as their control. In addition, their final analysis showed that treatment with isoniazid for 9 months, a currently recommended regimen for latent tuberculosis infection, did not decrease the risk for active tuberculosis (odds ratio, 0.94 [95% credible interval, 0.40 to 2.10]) compared with placebo. When faced with such inconsistencies, questions about the credibility of the network should be raised. Might the inclusion of poorer-quality trials have adversely affected the network? Many poor-quality trials escaped exclusion criteria (defined as high risk of bias across 4 categories) because bias was not able to be classified across 1 or more categories and was therefore deemed unclear. To their credit, the authors reported a subsequent analysis showing the absence of significant effect modification of trials with unclear or high risk of bias. However, because such a study may inform global policy, perhaps more stringent exclusion criteria should apply. These criteria may not only eliminate the inconsistencies listed here but also further reassure the reader about the equivalency between rifamycin-based regimens and isoniazid monotherapy. Reuben Arasaratnam, MD, MPH Baylor College of Medicine Houston, Texas Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOf InterestForms.do?msNum=L14-0554. Reference 1. Stagg HR, Zenner D, Harris RJ, Mun˜oz L, Lipman MC, Abubakar I. Treatment of latent tuberculosis infection: a network meta-analysis. Ann Intern Med. 2014;161:419-28. [PMID: 25111745] doi:10.7326/M14-1019
TO THE EDITOR: I read Stagg and colleagues' (1) review with
great interest. As far as I know, production of isolated isoniazid has been discontinued. However, the high prevalence of latent tuberculosis infection in many countries has necessitated the return of this preparation to the market because it is the only regimen for the management of this condition that is highly efficacious as well as cost-effective. Gauranga C. Dhar, MD Bangladesh Institute of Family Medicine and Research Dhaka, Bangladesh Disclosures: Authors have disclosed no conflicts of interest. Forms can be viewed at www.acponline.org/authors/icmje/ConflictOf InterestForms.do?msNum=L14-0555.
Reference 1. Stagg HR, Zenner D, Harris RJ, Mun˜oz L, Lipman MC, Abubakar I. Treatment of latent tuberculosis infection: a network meta-analysis. Ann Intern Med. 2014;161:419-28. [PMID: 25111745] doi:10.7326/M14-1019
IN RESPONSE: Dr. Arasaratnam raised several issues about
our systematic review, which was limited by the availability and quality of underlying articles because the literature for most latent tuberculosis infection regimens is relatively sparse. After rigorously assessing study quality, we tested for associations between quality indicators and outcomes but did not find particularly strong evidence for such associations; thus, this influence was likely minimal. Given the small number and frequently poor or unassessable quality of trials, further exclusion of studies would have limited, rather than provided assurance about, the comparison between rifamycins and isoniazid. Network meta-analyses pose a risk that indirect findings could be inconsistent with direct pairwise comparisons. We examined this possibility but did not find evidence to invalidate the network approach, although low statistical power was a limitation. To further validate our results, our parallel conventional meta-analysis showed similar effect estimates compared with the network analysis. Study quality would be of particular concern if it led to exaggeration of summary estimates (for example, an increase in efficacy). Varying study quality is likely to increase heterogeneity rather than induce bias in a particular direction. Substantial heterogeneity was observed in our study, leading to wider credible intervals and difficulties determining true differences in treatment effect. To some extent, we accounted for the heterogeneity by using a random-effects model. Despite the limitations of the included studies, evidence is available on the efficacy and safety of rifampicin regimens, including 3 to 4 months of monotherapy (1). Another recent meta-analysis also provides support for rifamycin-based regimens (2). Our systematic review supports the evidence that isoniazid for 6 months or longer is effective and safe. The observed limited efficacy of isoniazid monotherapy for 9 months is likely due to paucity of data. Previous analysis of trial data by Comstock suggests that the greatest efficacy of isoniazid regimens is likely achieved after 6 to 12 months of treatment (3). We agree with Dr. Dhar that isoniazid monotherapy remains an important treatment option to prevent progression from latent tuberculosis infection to active tuberculosis (4). The availability of appropriate isoniazid formulations therefore needs to be globally assured. The limitations outlined in our review indicate that further studies are needed to reliably evaluate the efficacy of treatment regimens for latent tuberculosis infection. Dominik Zenner, MD University College London and Public Health England London, United Kingdom Helen R. Stagg, PhD University College London London, United Kingdom
394 Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
www.annals.org
LETTERS Ross J. Harris, MSc Public Health England London, United Kingdom Ibrahim Abubakar, MBBS, PhD University College London and Public Health England London, United Kingdom Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=M14-1019.
Table. Correlation Between Confidence and Accuracy in Men and Women Confidence Rating
“I'm sure” “Feeling lucky” “No clue”
Confidence for All Questions Answered, %
Accuracy by Confidence Level, %
Men
Women
P Value*
Men
Women
P Value*
44.4 42.0 13.6
39.5 44.4 16.1
>0.001 0.180 0.27
78.3 49.8 32.0
80.5 53.5 32.0
>0.002 >0.001 1
* From Mann–Whitney U tests with Bonferroni adjustment.
References 1. A double-blind placebo-controlled clinical trial of three antituberculosis chemoprophylaxis regimens in patients with silicosis in Hong Kong. Hong Kong Chest Service/Tuberculosis Research Centre, Madras/British Medical Research Council. Am Rev Respir Dis. 1992;145:36-41. [PMID: 1731596] 2. Sharma SK, Sharma A, Kadhiravan T, Tharyan P. Rifamycins (rifampicin, rifabutin and rifapentine) compared to isoniazid for preventing tuberculosis in HIV-negative people at risk of active TB. Cochrane Database Syst Rev. 2013;7:CD007545. [PMID: 23828580] doi:10.1002/14651858.CD007545 .pub2 3. Comstock GW. How much isoniazid is needed for prevention of tuberculosis among immunocompetent adults? Int J Tuberc Lung Dis. 1999;3:847-50. [PMID: 10524579] 4. World Health Organization. Guidelines on the Management of Latent Tuberculosis Infection. Geneva: World Health Organization; 2014. Accessed at www.who.int/tb/publications/ltbi_document_page/en/ on 24 January 2015.
OBSERVATIONS The Association Between Confidence and Accuracy Among Users of a Mobile Web Platform for Medical Education Background: Recent literature suggests that physicians' diagnostic confidence tends to exceed accuracy, with confidence being relatively inflexible regardless of case difficulty (1). Overconfidence could lead to physicians not asking for help when they need it. Underconfidence may expose patients to defensive medicine. Those who lack confidence may misrepresent their actual level of knowledge, resulting in missed opportunities for professional development (2). Research suggests that female medical students perform as well as their male peers but report less confidence in their abilities and are actually perceived to be less confident (3). The future of medical education may involve training students to better align confidence and accuracy. One strategy may include providing students with immediate, actionable performance feedback through technology-enabled education platforms (4). Osmosis is a mobile Web platform designed to help medical students learn by answering questions related to their curricula. Its database includes information from more than 14 000 users. We developed a user-facing “Calibration Index” to show students their propensity for under- or overconfidence when answering questions. Objective: To measure medical students' confidence in their answers, compare confidence with accuracy, and examine differences between men and women. Methods: To answer a question in the platform, a user must first select 1 of 3 confidence ratings: “I'm sure,” “Feeling lucky,” or “No clue.” We analyzed the responses of 1021 users who answered more than 50 random multiple-choice and www.annals.org
true-or-false questions. An automated software package (SexMachine, version 0.1.1, based on Gender.c, version 1.2) that evaluates data on how frequently a name correlates with male or female gender in the United States was used to determine a user's sex from his or her first name. Users with ambiguous name–sex pairings were omitted. The final list was checked to confirm reasonable assignment. Users were assumed to be predominantly medical students in their preclinical years because the platform is designed for and marketed to this audience. Results are presented as the means of user-level response averages and compared with Mann–Whitney U tests using the Bonferroni adjustment (Stata, version 12.0, StataCorp). Findings: The study included 617 men and 404 women. Women's answers were more accurate than men's (61.4% vs. 60.3% correct; P = 0.040); however, they were less confident than men. For example, women selected “I'm sure” significantly less often than men (39.5% vs. 44.4% of responses) (Table). Average accuracy for questions rated as “I'm sure” and “Feeling lucky” was higher in women (80.5% vs. 78.3% of “I'm sure” responses were correct; 53.5% vs. 49.8% of “Feeling lucky” responses were correct). The average number of questions answered by female and male users did not differ (647 vs. 579; P = 0.40). When results were stratified into quartiles based on the number of questions answered, differences by sex remained. Discussion: The data suggest that, despite performing at the same level or higher, women lacked confidence relative to men. To understand the meaning and magnitude of this observed effect, it will be important to relate question-answering behavior to actual behavior on the wards. One limitation is that sex was determined from first name. Also, we could not verify that users were medical students or stratify results according to school and academic year. Notwithstanding these limitations, the data are a reminder that less confidence might not indicate a lack of knowledge and that confidence should not be mistaken for correctness. Insights gained from understanding the relationship between confidence and accuracy in medical trainees may ultimately prove valuable in reducing diagnostic errors caused by overconfidence and sex disparities caused by lack of confidence. Going forward, new educational interventions like the Calibration Index may help future physicians align confidence with accuracy to improve patient care and promote career advancement (5). Jason Theobald, BA New York University School of Medicine New York, New York Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015 395
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
LETTERS Shiv Gaglani, BA M. Ryan Haynes, BS, MPhil, PhD Johns Hopkins School of Medicine Baltimore, Maryland Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L14-0470. References 1. Meyer AN, Payne VL, Meeks DW, Rao R, Singh H. Physicians' diagnostic accuracy, confidence, and resource requests: a vignette study. JAMA Intern Med. 2013;173:1952-8. [PMID: 23979070] doi:10.1001/jamainternmed.2013 .10081 2. Nomura K, Yano E, Fukui T. Gender differences in clinical confidence: a nationwide survey of resident physicians in Japan. Acad Med. 2010;85:647-53. [PMID: 20354381] doi:10.1097/ACM.0b013e3181d2a796 3. Blanch DC, Hall JA, Roter DL, Frankel RM. Medical student gender and issues of confidence. Patient Educ Couns. 2008;72:374-81. [PMID: 18656322] doi:10.1016/j.pec.2008.05.021 4. Berner ES, Graber ML. Overconfidence as a cause of diagnostic error in medicine. Am J Med. 2008;121:S2-23. [PMID: 18440350] doi:10.1016 /j.amjmed.2008.01.001 5. Gaglani SM, Haynes MR. What can medical education learn from Facebook and Netflix? Ann Intern Med. 2014;160:640-1. [PMID: 24798524] doi:10.7326 /M13-2286
Progress and Challenges in Electronic Health Record Adoption: Findings From a National Survey of Physicians Background: The United States is modernizing the information technology infrastructure of its health care system. The provisions of the American Recovery and Reinvestment Act that are known as the Health Information Technology for Economic and Clinical Health (HITECH) Act provide substantial economic incentives for physicians to implement and use electronic health records (EHRs) (1, 2). Since the passage of the HITECH Act, studies have consistently found EHR use increasing steadily among office-based physicians (3, 4). However, recent studies report that a substantial proportion of physicians are unsure about or are not planning to participate in the program (5). Understanding which physicians may opt out will be necessary to achieve the goals of the HITECH Act and implement broader changes. Methods: We surveyed a sample of 3437 U.S. physicians selected to represent all primary care physicians and certain specialist physicians who were likely to care for a given patient over an extended time. We collected data in 2 waves. Wave 1 took place between October 2011 and March 2012, and wave 2 took place between May and July 2013. Detailed information about the survey methods is available from the authors on request. In this observation, we describe data from physicians who responded to both rounds of the survey (44% overall response rate). Our primary goal was to measure associations between the stage of EHR adoption and practice characteristics. Findings: In 2011, 44% of physicians had an EHR that met basic criteria (early adopters). Between 2011 and 2013, an additional 19% adopted a basic EHR (new adopters). In 2013, 20% were in the process of implementing an EHR or had implemented one without some functions required for a basic
EHR (partial implementers), 8% were planning to adopt an EHR in the next 2 years (planners), and 9% were not planning to adopt an EHR (persistent nonadopters). On average, persistent nonadopters were older than other physicians. The mean number of physicians employed in the main practice location of persistent nonadopters was 2.3 compared with 33.4 among early adopters and 15.1 among new adopters. Persistent nonadopters were significantly more likely to be employed in independent solo or 2-physician practices than early and new adopters, who were significantly more likely to be employed by a hospital or medical school, group or staff-model HMO, or network owned by a hospital or other type of health care organization. Methods of compensation varied by stage of implementation. Most persistent nonadopters reported fee-for-service as their primary compensation, whereas early and new adopters were more likely to report salary adjusted for performance. Persistent nonadopters seemed less likely to participate in incentive programs focused on improving the quality and continuity of care and were significantly less likely than early and new adopters to receive or have the potential to receive additional payments for managing patients with chronic conditions or complex needs. Discussion: Persistent nonadopters in small, isolated practices may be facing a unique set of challenges that limits their ability to adopt an EHR. Failure to address the needs of these physicians has implications beyond adoption because new models of health care delivery require the use of an EHR. Physicians who choose not to make the change to EHRs may find themselves further isolated if these new models become widespread, but they may move toward adoption as the penalty phase of the meaningful use program draws closer. If so, they are likely to require extensive support in selecting, implementing, and using these systems. Catherine DesRoches, PhD Mathematica Policy Research Cambridge, Massachusetts Disclosures: Disclosures can be viewed at www.acponline.org /authors/icmje/ConflictOfInterestForms.do?msNum=L14-0466. References 1. The American Recovery and Reinvestment Act, Pub L. No. 111-5, Stat. 123 (2009). 2. Blumenthal D. Wiring the health system— origins and provisions of a new federal program. N Engl J Med. 2011;365:2323-9. [PMID: 22168647] doi: 10.1056/NEJMsr1110507 3. Furukawa MF, King J, Patel V, Hsiao CJ, Adler-Milstein J, Jha AK. Despite substantial progress in EHR adoption, health information exchange and patient engagement remain low in office settings. Health Aff (Millwood). 2014; 33:1672-9. [PMID: 25104827] doi:10.1377/hlthaff.2014.0445 4. Audet AM, Squires D, Doty MM. Where are we on the diffusion curve? Trends and drivers of primary care physicians' use of health information technology. Health Serv Res. 2014;49:347-60. [PMID: 24358958] doi:10.1111 /1475-6773.12139 5. Hsiao C, Hing E. Use and Characteristics of Electronic Health Record Systems Among Office-based Physician Practices: United States, 2001-2013. NCHS Data Brief no. 143. Hyattsville, MD: National Center for Health Statistics; 2014.
396 Annals of Internal Medicine • Vol. 162 No. 5 • 3 March 2015
Downloaded From: https://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a Universite Laval Biblioteque User on 07/30/2017
www.annals.org
