Michael R. Jiroutek

Sulfonylurea Prescribing Patterns After the Introduction of DPP-4 Inhibitors and GLP-1 Receptor Agonists

PURPOSE Although newer agents (dipeptidyl peptidase [DPP]-4 inhibitors and glucagon-like peptide [GLP]-1 receptor agonists) are available for the treatment of hyperglycemia in patients with type 2 diabetes mellitus (T2DM), the impact of the availability of these agents on the use of second-generation sulfonylureas (SUs) is unknown. This article presents percentages of patients prescribed SUs, using data from the National Ambulatory Medical Care Survey (NAMCS). The associations between SU prescribing and prespecified variables of interest were also explored. METHODS The NAMCS database was queried for visits of patients aged ≥18 years with an International Classification of Diseases, Ninth Revision diagnostic code relevant to T2DM. χ(2) tests were conducted to assess the associations between SU use and year-group (2003-2004, 2007-2008, or 2009-2010) and other variables of interest. A multivariate logistic regression model was constructed to jointly assess the value of these variables in predicting SU use. All analyses were weighted using procedures recommended by the National Center for Health Statistics. FINDINGS Data from 7042 eligible visits were included, representing an extrapolated national estimate of 280,733,405 patient visits. The percentages of patients who received a prescription for an SU, by study year, were 25.7%, 23.4%, and 23.7% in 2003 to 2004, 2007 to 2008, and 2009 to 2010, respectively (P = 0.57). In the multivariate model, age ≥70 years, male sex, nonwhite race, primary care physician seen, and concurrent DPP-4 inhibitor use were significantly associated with SU use. IMPLICATIONS No significant decrease in the use of SUs was observed after the introduction of DPP-4 inhibitors and GLP-1 receptor agonists. However, patient-specific factors (eg, select demographic variables, site of care, and concurrent medication use) were associated with SU use.

Utilization of angiotensin converting enzyme inhibitors (ACEI) and angiotensin receptor blockers (ARB) in patients diagnosed with diabetes: Analysis from the National Ambulatory Medical Care Survey

Objective The objective of this study was to determine if a difference exists in the proportion of visits for the prescribing of angiotensin converting enzyme inhibitors (ACEI), or angiotensin receptor blockers (ARBs) in diabetic patients during 2007–2010. Methods This retrospective, cross-sectional, observational study included adults diagnosed with diabetes mellitus from the National Ambulatory Medical Care Survey (NAMCS) during 2007–2010. Weighted chi-square tests and a multivariable logistic regression model were used to analyze associations between ACEI/ARB prescriptions and predictors of interest. Odds ratios and 95% confidence intervals were reported. Results An unweighted total of 13,590 outpatient ambulatory care visits were identified for adult patients with diabetes without contraindications to ACEIs or ARBs in the NAMCS for the years studied. No statistically significant increase in the proportion of visits with an ACEI/ARB prescription was identified for years 2007–2010 (28.1% in 2007 to 32.2% in 2010). Females (OR 0.78, 95% CI 0.69- 0.89), patients 18–39 years old (OR 0.56, 95% CI 0.43- 0.75), and Medicare users (OR 0.81, 95% CI 0.70- 0.94) were significantly less likely to receive an ACEI/ARB prescription. Patients with hypertension (OR 2.80, 95% CI 2.39-3.29), hyperlipidemia (OR 1.42, 95% CI 1.22-1.65), and ischemic heart disease (OR 1.36, 95% CI 1.10-1.70) were significantly more likely to receive an ACEI/ARB prescription. Conclusions Despite extensive evidence showing the benefits of ACEI/ARB medications in diabetic patients, disparities of treatment remain evident.

Buying a significant result: Do we need to reconsider the role of the P value?

As readers of this journal are undoubtedly aware, publication bias, ie, the (positive) outcome of a study that influences the likelihood of publication, remains a significant issue in the research literature. Results with P values less than the “magical” .05 level and/or confidence intervals (CIs) that do not contain the null value of interest (and are hence deemed statistically significant) are much more predominant in published research literature than are studies without such findings. Studies in the field of hypertension are no exception to this bias. A relatively small cadre of researchers argue that, for a variety of reasons, nonstatistically significant results also need to be published. However, equally important is trusting that findings which are deemed statistically significant are truly important and not artefacts of statistical chicanery related to study design or analysis. This intellectual honesty remains a fundamental underpinning of the scientific method as it is known and applied today. Unfortunately, as a result of a perceived pressure (real or not) to achieve statistically significant results to increase the chances of publication, a variety of forms of ignorance (or worse, deception) still exist. In this article, we highlight one such issue that we feel is important to raise awareness among readers of this journal. Joining the cacophony of recommendations urging caution in the utilization and interpretation of P values,1-4 we recently recommended the use of CIs in addition to P values.5 The ability of the researcher (or reader) to gauge the magnitude and variability associated with the point estimate of the parameter under study remains a major added value of a CI and cannot be determined from a P value alone. However, for any such test statisticCI pairing, the two concepts are constructed from the same theoretical foundation and hence share many properties, both strengths and weaknesses. One major weakness is that, holding everything else constant, regardless of whether one is computing a test statistic with which to determine a P value or constructing a CI, a statistically significant result is more likely to be obtained as the sample size increases. This is an important issue—it allows a researcher to simply increase a study’s size to gain (or more cynically, “buy”) statistical significance. This editorial explains the origin of this issue, provides an illuminating example, and gives some suggestions as to how to avoid generating results whose significance may be attributed primarily to study size.

In Praise of Confidence Intervals: Much More Informative Than P Values Alone

Modern statistical inference approacheswere derived and published in the 1920s and 1930s by visionaries Sir Ronald Fisher, Jerzy Neyman, and Egon Pearson. While Fisher’s andNeyman-Pearson’s approach and solution to the problem differed, their methods have essentially blended over time into a single methodology that is used ubiquitously across science. As a result, the fundamental underpinnings on which hypothesis testing and confidence interval (CI) constructionwere built have remained in place for nearly a century. In multiple disciplines, a statistically significant P value obtained during hypothesis testing has for decades been deemed a validation of “successful” research. However, cautions in the use of hypothesis testing that rely solely on the interpretation of P values have increased in recent years. In the specific context of developing new drugs or new drug combination therapies for hypertension (or any other disease or condition of clinical concern), attainment of a statistically significant treatment effect alone is not enough to declare success: compelling evidence of clinical significance is also required. Such evidence is facilitated by the employment of CIs. This Editorial elucidates the connection between hypothesis tests and CIs, and explainswhy theuse ofCIs in place of a hypothesis-testing approach alone is preferred and encouraged by the authors.

Frequency of asthma education in primary care in the years 2007-2010.

Abstract Objective: Recent research suggests that health disparities persist among asthmatic patients and receipt of asthma education, though recent guidelines have highlighted the importance of receiving asthma education. The purpose of this study was to identify trends in the receipt of asthma education as well as to identify disparities in asthma education using the most recently available data in National Ambulatory Medical Care Survey, 2007–2010. Methods: Weighted chi-square tests were conducted to identify associations between asthma education and variables of interest. A weighted multivariate logistic regression model was subsequently constructed to jointly assess the association of factors of interest on receipt of asthma education. Submission to the Campbell University Institutional Review Board resulted in expedited approval. Results: The percentage of patients who receive asthma education remains quite low. After adjusting for all variables of interest: no statistically significant difference in receipt of asthma education between year groups (2007–2008, 2009–2010) was found (odds ratio [OR] 0.84, 95% confidence interval [CI] 0.52–1.34); patients seen by pediatricians (vs. internal medicine physicians) and Hispanic or Latino patients (vs. non-Hispanic or Latino patients) were more likely to receive asthma education (OR 2.72, 95% CI 1.11–6.66 and OR 2.33, 95% CI 1.18–4.60, respectively); and patients not prescribed a controller medication were less likely to receive asthma education than those who were (OR 0.56, 95% CI 0.37–0.82). Conclusions: Combined with previously published results, it appears the provision of asthma education continues to be low, despite proven benefits. Additionally, some patient and physician characteristics may be associated with the delivery of asthma education.

Why it is nonsensical to use retrospective power analyses to conduct a postmortem on your study

A postmortem (best known in the medical context) is strictly defined as an examination of a dead body to determine the cause of death. For some time, this same idea has been used by businesses to evaluate a failed project or venture, after the fact, in order to glean what went wrong and to avoid similar future mistakes.1 The obvious shortcoming of this paradigm is that once the project has failed, the “postmortem” findings cannot resuscitate the project, enabling it not to fail. In recent years, in an attempt to be more proactive in making such assessments, the idea of a “premortem” has taken hold in the business community.2 The term “premortem” describes the idea that, following the conception of a project, but before the actual start, the team brainstorms about what could cause the project to fail and makes any necessary adjustments in advance to avoid these potential pitfalls. Done thoroughly, this proactive strategy will make a project more robust, increasing the chances of its success. The business community seems to have finally figured out what the statistical/research community has known for decades—making an assessment of a project’s success in advance is far more useful than waiting until after it has failed to try to determine what went wrong. The statistical community has a wellknown calculation to help determine this: power. Statistical power is defined as the probability that the null hypothesis will be rejected, given that the alternative hypothesis is true. An a priori power calculation is a statistical analog to the premortem. Prestudy, we evaluate the likelihood of success by guesstimating known key factors, such as the effect size, sample size, data variability, and the desired type I and type II error rates. Unfortunately, the statistical/research community appears to be falling into the trap that plagued the business community—regressing to the use of the statistical analog of the postmortem, despite repeated warnings against the fallacy of attempting post hoc power calculations.3-12 Often referred to as retrospective power analysis, in this scenario, after a study has failed (operationally conceptualized in the statistical/research community as “failing to reject the null hypothesis”), a power calculation is undertaken to determine what sample size (or effect size) would have been needed to have been able to reject the null hypothesis; that is, to determine how the study could have been made successful after it died. To understand why such a calculation is not only useless, but also flat out nonsensical, consider the lottery. 2 | CONCEPTUAL ANALOGY

An analysis of insurance and other factors associated with asthma-related emergency department visits, 2009–2014

Disclosures: Authors have nothing to disclose. [2] Robinson D, Humbert M, Buhl R, et al. Revisiting type 2-high and type 2-low airway inflammation in asthma: current knowledge and therapeutic implications. Clin Exp Allergy. 2017;47:161e175. [3] National Institute for Health and Care Excellence. Measuring fractional exhaled nitric oxide concentration in asthma. NICE diagnostics guidance 12. www.nice. org.uk/dg12. Accessed August 24, 2017. [4] Kim SH, Moon JY, Kwak HJ, et al. Comparison of two exhaled nitric oxide analyzers: the NIOX MINO hand-held electrochemical analyzer and the NOA280i stationary chemiluminescence analyzer. Respirology. 2012;17:830e834. [5] Maniscalco M, de Laurentiis G, Weitzberg E, et al. Validation study of nasal nitric oxide measurements using a hand-held electrochemical analyser. Eur J Clin Invest. 2008;38:197e200. [6] Global strategy for asthma management and prevention 2015 (update). www. ginasthma.org. Accessed August 24, 2017. [7] American Thoracic Society, European Respiratory Society. ATS/ERS recommendations for standardized procedures for the online and offline measurement of exhaled lower respiratory nitric oxide and nasal oxide. Am J Respir Crit Care Med. 2005;171:912e930.

A Retrospective Cross-sectional Analysis of Health Education Disparities in Patients With Diabetes Using Data From the National Ambulatory Medical Care Survey

Purpose The purpose of this study was to determine if there was an association between receipt of diet/nutrition, exercise, and weight loss education in adult patients with a primary diagnosis of diabetes with various demographic and socioeconomic variables using data from the National Ambulatory Medical Care Survey (NAMCS) for the years 2008 to 2011. Methods This retrospective, cross-sectional, observational study design included patients ≥ 18 years of age with diabetes in the NAMCS between 2008 and 2011, inclusive. A series of weighted multivariable logistic regression models was constructed to evaluate predictors of diet/nutrition, exercise, and weight loss education. Odds ratios and 95% confidence intervals were reported. Results Among patients included in this study (n = 3027), 35.6% received diet/nutrition education, 21.8% received exercise education, and 13.6% received weight loss education. From the multivariable analyses, visits using “other” payment type, visits with Medicaid, and visits occurring in non-Metropolitan Statistical Areas were significantly less likely to receive diet/nutrition education; visits using other payment type, visits in non-Metropolitan Statistical Areas, and visits by those ≥ 65 and 45-64 years of age were significantly less likely to receive exercise education. No significant disparities in the receipt of weight loss education were found. Conclusion These findings indicate that although only approximately one third or fewer patients diagnosed with diabetes were receiving diet/nutrition, exercise, or weight loss education, there appeared to be limited disparities among the groups studied. Education rates appear to be trending upward over time, to be slightly improved as compared with previous studies, and to include fewer disparities.

Racial and Ethnic Disparities in Infant Mortality in North Carolina, 2008–2009

BACKGROUND This study was designed to investigate whether racial and ethnic disparities in infant mortality still exist in North Carolina and to examine predictors of infant mortality using the North Carolina Vital Statistics Dataverse. METHODS This was a retrospective, cross-sectional, observational study that included all 257,543 births in North Carolina in 2008–2009. Infant mortality was assessed based on birth records included in the database. Infant births and deaths were summarized by demographic and maternal/infant characteristics. A multivariate logistic regression model was constructed to jointly assess predictors of infant mortality. RESULTS The overall infant mortality rate in North Carolina was 0.8%. Adjusting for confounders through the construction and assessment of a multivariate logistic regression model, statistically significant associations were found between infant mortality and each of the following: maternal race (both black and ‘other’ versus white), infant sex, both premature and preterm gestation (versus full term), birth weight (both low and high versus normal), maternal education (both less than high school graduate and more than high school versus college graduate), prenatal care (both intermediate and inadequate versus adequate), and maternal tobacco use. LIMITATIONS Maternal race was limited to white, black, and other. Data on socioeconomic status, maternal medical risk factors, and quality of prenatal care were not available. At the time of the analysis, data for years beyond 2009 were limited. CONCLUSIONS Racial disparities in infant mortality persist in North Carolina; specifically, infants of nonwhite mothers have a higher mortality rate than do infants of white mothers. Other factors that continue to play a significant role in infant mortality in North Carolina include preterm and premature births, male infant sex, low birth weight, maternal education less than college graduate, maternal tobacco use, and less than adequate prenatal care.