Ten-Year Trends in Lithium Prescribing in Alberta, Canada.

AIMS: Despite lithium's clinical efficacy, it is commonly thought that its use is declining. The objective of this study is to describe the new and prevalent lithium users as well as rates of discontinuation of lithium use over a 10-year period. METHODS: This study used provincial administrative health data from Alberta, Canada between January 1, 2009 and December 31, 2018. Lithium prescriptions were identified within the Pharmaceutical Information Network database. Total and subgroup specific frequencies of new and prevalent lithium use were determined over the 10-year study period. Lithium discontinuation was also estimated through survival analysis. RESULTS: Between the calendar years of 2009 and 2018, 580,873 lithium prescriptions were dispensed in Alberta to 14,008 patients. The total number of new and prevalent lithium users appears to be decreasing over the 10-year timeframe, although the decline may have stopped or reversed in the latter years of the study period. Prevalent use of lithium was lowest among individuals between the ages of 18-24 years while the highest number of prevalent users were in the 50-64 age group, particularly among females. New lithium use was lowest amongst those 65 years and older. More than 60% (8,636) of patients prescribed lithium, discontinued use during the study timeframe. Lithium users between ages of 18-24 years were at the highest risk of discontinuations. CONCLUSIONS: Rather than a general decline in prescribing, trends in lithium use are dependent on age and sex. Further, the period soon after lithium initiation appears to be a key time period in which many lithium trials are abandoned. Detailed studies using primary data collection are needed to confirm and further explore these findings. These population-based results not only confirm a decline in lithium use, but also suggest that this may have stopped or even reversed. Population-based data on discontinuation pinpoint the period soon after initiation as the time when trials are most often discontinued.

Reporting of Model Performance and Statistical Methods in Studies That Use Machine Learning to Develop Clinical Prediction Models: Protocol for a Systematic Review.

BACKGROUND: With the growing excitement of the potential benefits of using machine learning and artificial intelligence in medicine, the number of published clinical prediction models that use these approaches has increased. However, there is evidence (albeit limited) that suggests that the reporting of machine learning-specific aspects in these studies is poor. Further, there are no reviews assessing the reporting quality or broadly accepted reporting guidelines for these aspects. OBJECTIVE: This paper presents the protocol for a systematic review that will assess the reporting quality of machine learning-specific aspects in studies that use machine learning to develop clinical prediction models. METHODS: We will include studies that use a supervised machine learning algorithm to develop a prediction model for use in clinical practice (ie, for diagnosis or prognosis of a condition or identification of candidates for health care interventions). We will search MEDLINE for studies published in 2019, pseudorandomly sort the records, and screen until we obtain 100 studies that meet our inclusion criteria. We will assess reporting quality with a novel checklist developed in parallel with this review, which includes content derived from existing reporting guidelines, textbooks, and consultations with experts. The checklist will cover 4 key areas where the reporting of machine learning studies is unique: modelling steps (order and data used for each step), model performance (eg, reporting the performance of each model compared), statistical methods (eg, describing the tuning approach), and presentation of models (eg, specifying the predictors that contributed to the final model). RESULTS: We completed data analysis in August 2021 and are writing the manuscript. We expect to submit the results to a peer-reviewed journal in early 2022. CONCLUSIONS: This review will contribute to more standardized and complete reporting in the field by identifying areas where reporting is poor and can be improved. TRIAL REGISTRATION: PROSPERO International Prospective Register of Systematic Reviews CRD42020206167; https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=206167. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR1-10.2196/30956.

Urine creatinine excretion and clinical outcomes in CKD.

BACKGROUND AND OBJECTIVES: Twenty-four-hour urine creatinine excretion is a reliable approximation of muscle mass. Whether changes in urine creatinine predict clinical outcomes in persons with CKD is unknown. This work studied the relationship between urine creatinine and patient and renal survival in people with CKD not requiring renal replacement therapy. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: This longitudinal cohort study included incident stages 3-5 CKD patients referred to the renal clinic at the University Federico II in Naples between January of 1995 and December of 2005. Clinical data and urine creatinine were updated at each visit. Main outcomes were all-cause mortality and kidney failure requiring dialysis. RESULTS: This study enrolled 525 individuals and followed them for a median of 6 years (range of 4 months to 15 years). Urine creatinine excretion declined by 16 mg/d per year (95% confidence interval, 14 to 19) in participants with CKD stages 3a, 3b, and 4, and it remained stable in participants with stage 5 CKD. Per each 20 mg/d decline in urine creatinine, mortality increased by 3% (adjusted hazard ratio, 1.03; 95% confidence interval, 1.01 to 1.05), and the risk of initiating dialysis increased by 2% (adjusted hazard ratio, 1.02; 95% confidence interval, 1.01 to 1.03). These associations were independent of body mass index and GFR. CONCLUSIONS: In persons with CKD stages 3 and 4, urine creatinine declines at a rate of 16 mg/d per year. Lower urine creatinine excretion predicts greater risk of kidney failure and patient mortality.