The Unseen Hand: AI-Based Prescribing Decision Support Tools and the Evaluation of Drug Safety and Effectiveness.

The use of artificial intelligence (AI)-based tools to guide prescribing decisions is full of promise and may enhance patient outcomes. These tools can perform actions such as choosing the 'safest' medication, choosing between competing medications, promoting de-prescribing or even predicting non-adherence. These tools can exist in a variety of formats; for example, they may be directly integrated into electronic medical records or they may exist in a stand-alone website accessible by a web browser. One potential impact of these tools is that they could manipulate our understanding of the benefit-risk of medicines in the real world. Currently, the benefit risk of approved medications is assessed according to carefully planned agreements covering spontaneous reporting systems and planned surveillance studies. But AI-based tools may limit or even block prescription to high-risk patients or prevent off-label use. The uptake and temporal availability of these tools may be uneven across healthcare systems and geographies, creating artefacts in data that are difficult to account for. It is also hard to estimate the 'true impact' that a tool had on a prescribing decision. International borders may also be highly porous to these tools, especially in cases where tools are available over the web. These tools already exist, and their use is likely to increase in the coming years. How they can be accounted for in benefit-risk decisions is yet to be seen.

Incident gout and chronic Kidney Disease: healthcare utilization and survival.

BACKGROUND: Uncontrolled gout can cause significant joint and organ damage and has been associated with impairments in quality of life and high economic cost. Gout has also been associated with other comorbid diseases, such as chronic kidney disease. The current study explored if healthcare resource utilization (HRU) and survival differs between patients with incident gout in the presence or absence of chronic kidney disease (CKD). METHODS: Clalit Health Services (CHS) data were used to conduct a retrospective population-based cohort study of incident gout between 1/1/2006-31/12/2009. Incident cases of gout were identified and stratified by CKD status and by age group (< 55 and 55+ years). CKD status was defined as a pre-existing diagnosis of chronic kidney disease, chronic renal failure, kidney transplantation, or dialysis at index date. Demographic and clinical characteristics, as well as healthcare resource use, were reported. RESULTS: A total of 12,940 incident adult gout patients, with (n = 8286) and without (n = 4654) CKD, were followed for 55,206 person-years. Higher rates of HRU were observed for gout patients with CKD than without. Total annual hospital admissions for patients with gout and CKD were at least 3 times higher for adults < 55 (mean = 0.51 vs 0.13) and approximately 1.5 times higher for adults 55+ (mean = 0.46 vs 0.29) without CKD. Healthcare utilization rates from year 1 to year 5 remained similar for gout patients < 55 years irrespective of CKD status, however varied according to healthcare utilization by CKD status for gout patients 55+ years. The 5-year all-cause mortality was higher among those with CKD compared to those without CKD for both age groups (HR< 55 years = 1.65; 95% CI 1.01-2.71; HR55+ years = 1.50; 95% CI 1.37-1.65). CONCLUSIONS: The current study suggests important differences exist in patient characteristics and outcomes among patients with gout and CKD. Healthcare utilization differed between sub-populations, age and comorbidities, over the study period and the 5-year mortality risk was higher for gout patients with CKD, regardless of age. Future work should explore factors associated with these outcomes and barriers to gout control in order to enhance patient management among this high-risk subgroup.

Development and validation of algorithms to identify newly diagnosed type 1 and type 2 diabetes in pediatric population using electronic medical records and claims data.

PURPOSE: To develop and validate algorithms to classify diabetes type in newly diagnosed pediatric patients with DM. METHOD: Data from the United States Department of Defense health system were used to identify patients aged 10 to 18 years with incident DM. Two independent sets of 200 children were randomly sampled for algorithm development and validation. Algorithms were developed based on clinical insight, published literature, and quantitative approaches. The actual DM type was ascertained via chart review. Finally, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were evaluated. RESULTS: Among the 400 patients, mean age was 14.2 (±2.5 years), and 50% were female. The best performing algorithms were based on data available in claims. They consisted of several logical expressions based on one predictor or more, which classified patients by use of glucose-lowering drugs or testing, DM ICD-9 diagnosis codes, and comorbidities. The best performing T2DM and T1DM algorithms achieved 90% and 98% sensitivity, 95% and 95% specificity, 87% and 98% PPV, and 96% and 96% NPV, respectively. CONCLUSIONS: Our results suggest that claims algorithms can accurately identify newly diagnosed T1DM and T2DM pediatric patients, which can facilitate large database studies in children with T1DM and T2DM. However, external validation in other data sources is needed.

Real-world chart review study of adverse events management in patients taking tyrosine kinase inhibitors to treat metastatic renal cell carcinoma.

PURPOSE: The purpose is to describe management of adverse events of special interest across tyrosine kinase inhibitors approved for metastatic renal cell carcinoma. METHODS: We conducted a retrospective chart review in metastatic renal cell carcinoma patients initiating tyrosine kinase inhibitor monotherapy between 15 November 2010 and 15 November 2013, and experiencing ≥ 1 adverse events of special interest (diarrhea, fatigue, hand-foot syndrome, hypertension, or stomatitis/mucositis) within 3 months of initiation. Demographics, medical history, treatment regimens, and adverse events of special interest management data for 3.5 months postonset were collected. RESULTS: In 220 charts from 27 centers, tyrosine kinase inhibitors prescribed included sunitinib (55%), pazopanib (27%), axitinib (9%), and sorafenib (8%). During the study period, patients experienced 376 adverse events of special interest (13% serious). Fatigue was most common (62% of patients), followed by hypertension (37%), diarrhea (30%), stomatitis/mucositis (29%), and hand-foot syndrome (12%). Over half (56%) the adverse events of special interest were resolved or resolving. Treatment discontinuation due to adverse events of special interest occurred in 15% of patients. Prophylaxis was rarely provided (8%), whereas 59% of patients received adverse events of special interest treatment (pharmacologic (55%) and/or nonpharmacologic (7%)), 27% received tyrosine kinase inhibitor dose management, 23% received both adverse events of special interest treatment and dose management, and 31% received neither. Hypertension was the most treated (72% of all events), and fatigue was the least treated (9%); only 4% of patients received pharmacologic treatment for fatigue. CONCLUSIONS: Most adverse events of special interest were nonserious and more than half of the patients received pharmacologic and/or nonpharmacologic treatment. Fatigue was the most common yet least frequently treated adverse event of special interest, and few patients received prophylaxis or nonpharmacologic treatment. More emphasis on treatment and prophylaxis options for bothersome adverse events is warranted.

Clinical characteristics and healthcare utilization of patients with multicentric Castleman disease.

Multicentric Castleman disease (MCD) is a rare lymphoproliferative disease. Little is known about how patient clinical features and healthcare utilization varies by human immunodeficiency virus (HIV) status and disease subtype. Data of MCD patients identified between 2000 and 2009 were collected from medical records at two United States treatment centres. Clinical, demographic, and biochemical characteristics, drug therapies and medical utilization were descriptively reported by HIV status and cell histology, and statistically compared with the Fisher's Exact and Kruskal-Wallis tests. Patients (n = 59) had a pathologically and clinically confirmed MCD diagnosis: plasmacytic (42%), hyaline vascular (29%) and mixed (15%); 10% had HIV infection. In the first year after diagnosis, MCD patients on average saw a healthcare provider more than six times, were hospitalized at least once and underwent frequent radiological and laboratory testing. Rituximab was the most commonly used drug therapy, followed by corticosteroids and conventional chemotherapy. One- and 2-year survival was excellent in HIV-negative patients (100% and 97%, respectively) but inferior for HIV-positive cases (67% and 67%, respectively). Heterogeneous treatment decisions were observed in this MCD study; HIV status was the only distinguishing clinical criteria associated with pharmacotherapies. Additional research is necessary to guide treatment of this rare lymphoproliferative disorder.

The safety of bendamustine in patients with chronic lymphocytic leukemia or non-Hodgkin lymphoma and concomitant renal impairment: a retrospective electronic medical record database analysis.

Abstract This retrospective study compared adverse-event rates in patients with chronic lymphocytic leukemia (CLL) or non-Hodgkin lymphoma (NHL), with and without renal impairment, receiving bendamustine alone or with rituximab. Patients (n = 940) were stratified into a renally impaired group (creatinine clearance [CrCL] < 40 mL/min) and two comparator groups (CrCL ≥ 40 mL/min and CrCL ≥ 60 mL/min). Renally impaired patients with NHL had a significantly greater incidence of grade 3-4 thrombocytopenia compared with the CrCL ≥ 60 mL/min group (hazard ratio [HR], 2.57; p = 0.025). For CLL and NHL together, grade 3-4 increased blood urea nitrogen was significantly higher in the renally impaired group than in the CrCL ≥ 40 mL/min (HR, 2.36; p = 0.02) and CrCL ≥ 60 mL/min (HR, 4.46; p = 0.001) groups. Based on these results, monitoring blood counts (including platelets) and renal function would be prudent in the management of patients with renal dysfunction and NHL or CLL who receive bendamustine-based regimens.

Infection acquisition following intensive care unit room privatization.

BACKGROUND: Patients in intensive care units (ICUs) often acquire infections, which impose a heavy human and financial burden. The use of private rooms may reduce the acquisition of certain pathogens, but the limited evidence on this topic is inconsistent. METHODS: We compared the rates of acquisition of infectious organisms in an ICU before and after a change from multibed to single rooms. As a control, we used acquisition rates in the ICU of a nearby university teaching hospital, which contained both multibed and single rooms, during the study period. We used a statistical model to adjust for background time trends common to both hospitals. RESULTS: The adjusted rate of acquisition of Clostridium difficile, vancomycin-resistant Enterococcus species, and methicillin-resistant Staphylococcus aureus combined decreased by 54% (95% confidence interval [CI], 29%-70%) following the intervention. The methicillin-resistant S aureus acquisition rate fell by 47% (95% CI,1%-71%), the C difficile acquisition rate fell by 43% (95% CI, 7%-65%), and the yeast acquisition rate fell by 51% (95% CI, 34%-64%). Twelve common and likely exogenous organisms and exogenous/endogenous organisms had a reduction in acquisition rates after the intervention; for 6 of them, this reduction was statistically significant. No effect was observed on the acquisition rate of coagulase-negative Staphylococcus species, the most common endogenous organism, for which no change would be expected. The adjusted rate ratio of the average length of stay in the ICU was 10% (95% CI, 0%-19%) lower after the intervention. CONCLUSION: Conversion to single rooms can substantially reduce the rate at which patients acquire infectious organisms while in the ICU.