Multilevel Intervention for Low-Income Maternal Smokers in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC).

Objectives. To test the efficacy of Babies Living Safe and Smokefree (BLiSS), a multilevel intervention initiated in a citywide safety net health system to improve low-income maternal smokers' abstinence and reduce child tobacco smoke exposure. Methods. This randomized controlled trial in Philadelphia, Pennsylvania (2015-2020), recruited low-income maternal smokers who received a brief smoking intervention (Ask, Advise, Refer [AAR]) from nutrition professionals in the Special Supplemental Nutrition Program for Women, Infants, and Children before randomization to (1) a multilevel intervention (AAR + multimodal behavioral intervention [MBI]; n = 199) or (2) an attention control intervention (AAR + control; n = 197). Results. AAR + MBI mothers had significantly higher 12-month bioverified abstinence rates than did AAR + control mothers (odds ratio [OR] = 9.55; 95% confidence interval [CI] = 1.54, 59.30; P = .015). There were significant effects of time (b = -0.15; SE = 0.04; P < .001) and condition by time (b = -0.19; SE = 0.06; P < .001) on reported child exposure favoring AAR + MBI, but no group difference in child cotinine. Presence of other residential smokers was related to higher exposure. Higher baseline nicotine dependence was related to higher child exposure and lower abstinence likelihood at follow-up. Conclusions. The multilevel BLiSS intervention was acceptable and efficacious in a population that experiences elevated challenges with cessation. Public Health Implications. BLiSS is a translatable intervention model that can successfully improve efforts to address the persistent tobacco-related burdens in low-income communities. Trial Registration. Clinical Trials.gov identifier: NCT02602288. (Am J Public Health. 2022;112(3):472-481. https://doi.org/10.2105/AJPH.2021.306601).

Statistical inference for natural language processing algorithms with a demonstration using type 2 diabetes prediction from electronic health record notes.

The pointwise mutual information statistic (PMI), which measures how often two words occur together in a document corpus, is a cornerstone of recently proposed popular natural language processing algorithms such as word2vec. PMI and word2vec reveal semantic relationships between words and can be helpful in a range of applications such as document indexing, topic analysis, or document categorization. We use probability theory to demonstrate the relationship between PMI and word2vec. We use the theoretical results to demonstrate how the PMI can be modeled and estimated in a simple and straight forward manner. We further describe how one can obtain standard error estimates that account for within-patient clustering that arises from patterns of repeated words within a patient's health record due to a unique health history. We then demonstrate the usefulness of PMI on the problem of predictive identification of disease from free text notes of electronic health records. Specifically, we use our methods to distinguish those with and without type 2 diabetes mellitus in electronic health record free text data using over 400 000 clinical notes from an academic medical center.

Randomized Noninferiority Trial of Telephone vs In-Person Disclosure of Germline Cancer Genetic Test Results.

Background: Germline genetic testing is standard practice in oncology. Outcomes of telephone disclosure of a wide range of cancer genetic test results, including multigene panel testing (MGPT) are unknown. Methods: Patients undergoing cancer genetic testing were recruited to a multicenter, randomized, noninferiority trial (NCT01736345) comparing telephone disclosure (TD) of genetic test results with usual care, in-person disclosure (IPD) after tiered-binned in-person pretest counseling. Primary noninferiority outcomes included change in knowledge, state anxiety, and general anxiety. Secondary outcomes included cancer-specific distress, depression, uncertainty, satisfaction, and screening and risk-reducing surgery intentions. To declare noninferiority, we calculated the 98.3% one-sided confidence interval of the standardized effect; t tests were used for secondary subgroup analyses. Only noninferiority tests were one-sided, others were two-sided. Results: A total of 1178 patients enrolled in the study. Two hundred eight (17.7%) participants declined random assignment due to a preference for in-person disclosure; 473 participants were randomly assigned to TD and 497 to IPD; 291 (30.0%) had MGPT. TD was noninferior to IPD for general and state anxiety and all secondary outcomes immediately postdisclosure. TD did not meet the noninferiority threshold for knowledge in the primary analysis, but it did meet the threshold in the multiple imputation analysis. In secondary analyses, there were no statistically significant differences between arms in screening and risk-reducing surgery intentions, and no statistically significant differences in outcomes by arm among those who had MGPT. In subgroup analyses, patients with a positive result had statistically significantly greater decreases in general anxiety with telephone disclosure (TD -0.37 vs IPD +0.87, P = .02). Conclusions: Even in the era of multigene panel testing, these data suggest that telephone disclosure of cancer genetic test results is as an alternative to in-person disclosure for interested patients after in-person pretest counseling with a genetic counselor.

Comparative Effectiveness and Risks of Bowel Preparation Before Elective Colorectal Surgery.

OBJECTIVE: The objective of this study was to determine the relationship between bowel preparation and surgical site infections (SSIs), and also other postoperative complications, after elective colorectal surgery. BACKGROUND: SSI is a major source of postoperative morbidity/costs after colorectal surgery. The value of preoperative bowel preparation to prevent SSI remains controversial. METHODS: We analyzed 32,359 patients who underwent elective colorectal resections in the American College of Surgeons National Surgery Quality Improvement Program database from 2012 to 2014. Univariable and multivariable analyses were performed; propensity adjustment using patient/procedure characteristics was used to account for nonrandom receipt of bowel preparation. RESULTS: 26.7%, 36.6%, 3.8%, and 32.9% of patients received no bowel preparation, mechanical bowel preparation (MBP), oral antibiotics (OA), and MBP + OA, respectively. After propensity adjustment, MBP was not associated with decreased risk of SSI compared with no bowel preparation. In contrast, both OA and OA + MBP were associated with decreased risk of any SSI (adjusted odds ratio 0.49, 95% confidence interval 0.38-0.64; and adjusted odds ratio 0.45, 95% confidence interval 0.40-0.50, respectively) compared with no bowel preparation. OA and MBP + OA were associated with decreased risks of anastomotic leak, postoperative ileus, readmission, and also shorter length of stay (all P < 0.05). Bowel preparation was not associated with increased risk of cardiac/renal complications compared with no preparation. CONCLUSIONS: The use of MBP alone before elective colorectal resection to prevent SSI is ineffective and should be abandoned. In contrast, OA and MBP + OA are associated with decreased risks of SSI and are not associated with increased risks of other adverse outcomes compared with no preparation. Prospective studies to determine the efficacy of OA are warranted; in the interim, MBP + OA should be used routinely before elective colorectal resection to prevent SSI.

Relationships Among Financial Distress, Emotional Distress, and Overall Distress in Insured Patients With Cancer.

PURPOSE: Recent studies have demonstrated increasing rates of financial toxicities and emotional distress related to cancer treatment. This study assessed and characterized the relationships among financial distress, emotional symptoms, and overall distress in patients with cancer. METHODS: A cross-sectional sample of patients with cancer who visited our outpatient medical oncology and psychiatry clinics completed a pen-and-paper survey. The survey assessed demographics; cost concerns; and financial, emotional, and overall distress. RESULTS: One hundred twenty insured patients completed the survey. Sixty-five percent reported clinically significant overall distress scores, with the same percentage reporting at least one emotional problem (worry, anxiety, depression, etc). Twenty-nine percent scored in the range of high to overwhelming financial distress. By using structural equation modeling, we found that financial distress was associated with overall distress. This association was both direct (accounting for 76% of the effect) and indirect (accounting for 24% of the effect) via mediation by emotional distress. CONCLUSION: This cohort of patients with cancer reported significant levels of emotional distress, financial distress, and overall distress. These factors were interrelated, with both financial and emotional distress contributing to overall distress. Interventions targeted at alleviating financial distress may help to decrease levels of overall distress.

Survival Impact of Increasing Time to Treatment Initiation for Patients With Head and Neck Cancer in the United States.

PURPOSE: To estimate the overall survival (OS) impact from increasing time to treatment initiation (TTI) for patients with head and neck squamous cell carcinoma (HNSCC). METHODS: Using the National Cancer Data Base (NCDB), we examined patients who received curative therapy for the following sites: oral tongue, oropharynx, larynx, and hypopharynx. TTI was the number of days from diagnosis to initiation of curative treatment. The effect of TTI on OS was determined by using Cox regression models (MVA). Recursive partitioning analysis (RPA) identified TTI thresholds via conditional inference trees to estimate the greatest differences in OS on the basis of randomly selected training and validation sets, and repeated this 1,000 times to ensure robustness of TTI thresholds. RESULTS: A total of 51,655 patients were included. On MVA, TTI of 61 to 90 days versus less than 30 days (hazard ratio [HR], 1.13; 95% CI, 1.08 to 1.19) independently increased mortality risk. TTI of 67 days appeared as the optimal threshold on the training RPA, statistical significance was confirmed in the validation set (P < .001), and the 67-day TTI was the optimal threshold in 54% of repeated simulations. Overall, 96% of simulations validated two optimal TTI thresholds, with ranges of 46 to 52 days and 62 to 67 days. The median OS for TTI of 46 to 52 days or fewer versus 53 to 67 days versus greater than 67 days was 71.9 months (95% CI, 70.3 to 73.5 months) versus 61 months (95% CI, 57 to 66.1 months) versus 46.6 months (95% CI, 42.8 to 50.7 months), respectively (P < .001). In the most recent year with available data (2011), 25% of patients had TTI of greater than 46 days. CONCLUSION: TTI independently affects survival. One in four patients experienced treatment delay. TTI of greater than 46 to 52 days introduced an increased risk of death that was most consistently detrimental beyond 60 days. Prolonged TTI is currently affecting survival.

Why Summary Comorbidity Measures Such As the Charlson Comorbidity Index and Elixhauser Score Work.

BACKGROUND: Comorbidity adjustment is an important component of health services research and clinical prognosis. When adjusting for comorbidities in statistical models, researchers can include comorbidities individually or through the use of summary measures such as the Charlson Comorbidity Index or Elixhauser score. We examined the conditions under which individual versus summary measures are most appropriate. METHODS: We provide an analytic proof of the utility of comorbidity summary measures when used in place of individual comorbidities. We compared the use of the Charlson and Elixhauser scores versus individual comorbidities in prognostic models using a SEER-Medicare data example. We examined the ability of summary comorbidity measures to adjust for confounding using simulations. RESULTS: We devised a mathematical proof that found that the comorbidity summary measures are appropriate prognostic or adjustment mechanisms in survival analyses. Once one knows the comorbidity score, no other information about the comorbidity variables used to create the score is generally needed. Our data example and simulations largely confirmed this finding. CONCLUSIONS: Summary comorbidity measures, such as the Charlson Comorbidity Index and Elixhauser scores, are commonly used for clinical prognosis and comorbidity adjustment. We have provided a theoretical justification that validates the use of such scores under many conditions. Our simulations generally confirm the utility of the summary comorbidity measures as substitutes for use of the individual comorbidity variables in health services research. One caveat is that a summary measure may only be as good as the variables used to create it.