Rethinking PICO in the Machine Learning Era: ML-PICO

Background  Machine learning (ML) has captured the attention of many clinicians who may not have formal training in this area but are otherwise increasingly exposed to ML literature that may be relevant to their clinical specialties. ML papers that follow an outcomes-based research format can be assessed using clinical research appraisal frameworks such as PICO (Population, Intervention, Comparison, Outcome). However, the PICO frameworks strain when applied to ML papers that create new ML models, which are akin to diagnostic tests. There is a need for a new framework to help assess such papers. Objective  We propose a new framework to help clinicians systematically read and evaluate medical ML papers whose aim is to create a new ML model: ML-PICO (Machine Learning, Population, Identification, Crosscheck, Outcomes). We describe how the ML-PICO framework can be applied toward appraising literature describing ML models for health care. Conclusion  The relevance of ML to practitioners of clinical medicine is steadily increasing with a growing body of literature. Therefore, it is increasingly important for clinicians to be familiar with how to assess and best utilize these tools. In this paper we have described a practical framework on how to read ML papers that create a new ML model (or diagnostic test): ML-PICO. We hope that this can be used by clinicians to better evaluate the quality and utility of ML papers.     

Pharmacy patron perspectives of community pharmacist administered influenza vaccinations

Background

One approach to boost influenza vaccination coverage has been to expand immunization authority. In 2012, the province of Ontario gave community pharmacists the authority to administer the influenza vaccine.

Changes in the SF-12 among Depressed Elders Six Months after Discharge from an Inpatient Geropsychiatric Unit

Using the SF-12 to measure physical and mental functioning, the authors examine the intra-individual changes in health-related quality of life (HRQOL) 6 months post-discharge for depressed older adults. In addition, they examine three sets of predictors that might influence these changes. The sample of depressed older adults was recruited from an inpatient geropsychiatry unit. Although their physical and mental health scores on the SF-12 were lower than comparable norms, the sample showed an average increase in their mental functioning but a decrease in the physical functioning over the 6 months. Negative life-events were significant predictors of people who reported no change in their mental health functioning and decreases in their physical health functioning. Interestingly, those who experienced positive life events were more likely to report declines and younger participants were more likely to report no change in their physical functioning. The findings indicate that the effects of depression on HRQOL can have enduring effects on a sample of previously hospitalized older adults. The significance of life event changes might signify the importance of taking into account non-traditional areas of medical interventions. Further, the findings indicate the usefulness of the SF-12 quantifying HRQOL outcomes.     

Alcohol consumption patterns and predictors of use following liver transplantation for alcoholic liver disease.

For patients who receive a liver transplant (LTX) for alcoholic liver disease (ALD), investigators are focusing beyond survival to determine specific alcohol use outcomes. Studies suggest the use of alcohol ranges from 8 to 22% for the first post-transplant year with cumulative rates reaching 30 to 40% by 5 years following transplantation. Yet while investigators are interested in determining specific rates of alcohol use and predictors of use, only three studies since 1990 have been prospective. In 1998, we began a prospective study of post-LTX alcohol consumption in ALD recipients using multiple repeated measures of alcohol use. After 5 years of follow-up, we found that 22% had used any alcohol by the first year and 42% had a drink by 5 years. By 5 years, 26% drank at a heavier use (binge) pattern and 20% drank in a frequent pattern. In a univariate model, predictors of alcohol use included pre-transplant length of sobriety, a diagnosis of alcohol dependence, a history of other substance use, and prior alcohol rehabilitation.     

Reduction in oral corticosteroid use with mometasone furoate dry powder inhaler improves health-related quality of life in patients with severe persistent asthma.

Severe persistent asthma can have a substantial impact on a patient's health-related quality of life (HRQL), both as a result of symptoms and from side effects of treatment. The HRQL impact of two doses (400 and 800 microg twice daily) of mometasone furoate dry powder inhaler (MF DPI) was compared with placebo in patients with severe persistent asthma previously maintained on oral steroids as a component of a previously published randomized, 12-week, double-blind, placebo-controlled, multicenter trial. A 9-month open-label extension (OLE), with all patients treated with MF DPI, followed. Patients 12 years of age or older completed a generic HRQL measure, the Medical Outcomes Trust Short Form-36 (SF-36), and an asthma-specific measure, the Marks Asthma Quality of Life Questionnaire (AQLQ-M), at baseline, at endpoint (last evaluable visit) of the double-blind phase (EODBP), and after the first 3 months of the OLE. Of 132 patients enrolled in the study, 128 provided HRQL data at baseline and at EODBP. Mean SF-36 scores at baseline showed significant HRQL impairment compared with U.S. general population norms. With treatment, the reduction in oral corticosteroid (OCS) requirements of the MF-DPI-treated groups was accompanied by significant (p < 0.05) improvement over placebo in the physical domain of HRQL (SF-36 physical component summary score and the physical function subscale) at EODBP. MF-DPI-treated patients also showed significant improvements at EODBP in each of the four subscales of the AQLQ-M (p<0.05). From EODBP to the OLE 3-month endpoint, patients treated with MF DPI twice daily maintained, or improved, SF-36 scores in most domains. Symptomatic improvement and reduction in OCS use with MF DPI were accompanied by significant improvement in HRQL in patients with severe persistent asthma. These improvements were maintained during the 3-month period of the OLE in which HRQL was evaluated.