Schapira, Marilyn M.

AbstractBackgroundTo determine the impact of personalized risk communication and opioid prescribing on nonprescribed opioid use, we conducted a secondary analysis of randomized controlled trial participants followed prospectively for 90 days after an emergency department (ED) visit for acute back or kidney stone pain.MethodsA total of 1301 individuals were randomized during an encounter at four academic EDs into a probabilistic risk tool (PRT) arm, a narrative‐enhanced PRT arm, or a general risk information arm (control). In this secondary analysis, both risk tool arms were combined and compared with the control arm. We used logistic regressions to determine associations between receiving personalized risk information, receiving an opioid prescription in the ED, and nonprescribed opioid use in general and by race.ResultsComplete follow‐up data were available for 851 participants; 23.3% (n = 198) were prescribed opioids (34.2% of White vs. 11.6% of Black participants, p < 0.001). Fifty‐six (6.6%) participants used nonprescribed opioids. Participants in the personalized risk communication arms had lower nonprescribed opioid use odds (adjusted odds ratio [aOR] 0.58, 95% confidence interval [CI] 0.4–0.83). Black versus White participants had greater nonprescribed opioid use odds (aOR 3.47, 95% CI 2.05–5.87, p < 0.001). Black participants who were prescribed opioids had a lower marginal probability of using nonprescribed opioids versus those who were not (0.06, 95% CI 0.04–0.08, p < 0.001 vs. 0.10, 95% CI 0.08–0.11, p < 0.001). The absolute risk difference in nonprescribed opioid use for Black and White participants, respectively, in the risk communication versus the control arm, was 9.7% and 0.1% (relative risk ratio 0.43 vs. 0.95).ConclusionsAmong Black but not White participants, personalized opioid risk communication and opioid prescribing were associated with lower odds of nonprescribed opioid use. Our findings suggest that racial disparities in opioid prescribing—which have been previously described within the context of this trial—may paradoxically increase nonprescribed opioid use. Personalized risk communication may effectively reduce nonprescribed opioid use, and future research should be designed specifically to explore this possibility in a larger cohort.

Objective The aim of the study was to determine how frequently mental status and mental status changes are documented in the written patient summary (“sign-out”) provided to covering physicians. Patients and Methods This was a retrospective cohort study of general medical patients hospitalized between March 16, 2009, and March 15, 2010, conducted at 2 teaching hospitals. Participants included patients with mental status change adverse events (MSAEs) and their providers. Chart review was performed to identify patients with MSAEs and details about these events. Sign-outs were reviewed for documentation of mental status. Main outcome measures were (1) proportion of patients with MSAEs who had mental status ever recorded in sign-out entries and (2) the proportion of patients with MSAEs whose change in mental status was recorded in the sign-out. Results Sixty-eight patients had MSAEs and were included in the sample. Fifty percent of MSAEs were attributed to medications; 75% of these events were first detected by nurses. Only 25% of patients with MSAEs had their change in mental status recorded in sign-outs. Conclusions Recording mental status in written sign-outs is uncommon. Particularly concerning is that patients with MSAEs identified by chart review seldom had sign-outs that reflected those events. Interventions should be designed to increase the recording of this information in sign-outs.

BACKGROUNDInaccurate or incomplete information in the written portion of the patient handoff, or sign‐out, may be associated with adverse events in hospitalized patients. Little is known about what information providers actually include in written sign‐out documents and how sign‐outs change over time.OBJECTIVES(1) Provide a descriptive analysis of initial and subsequent hospital day‐written sign‐out content, and (2) evaluate the relationship between team workload and sign‐out composition.DESIGNRetrospective review of sign‐out documents from a larger observational study of general medicine patients admitted to housestaff and hospitalist teams at 3 hospitals.MAIN MEASURESThe presence of 13 components of a high‐quality sign‐out. We performed descriptive analyses and compared initial and subsequent day sign‐outs for content.KEY RESULTSWe reviewed 200 patient hospitalizations (200 initial handoffs, 580 subsequent day handoffs). Initial sign‐out entries contained a mean of 7.54 (standard deviation: 2.27) key sign‐out components. Subsequent day sign‐outs contained a higher percentage of certain key elements but had more vague language. The number of elements present in the sign‐out was reduced as patient census increased (r = −0.295, P < 0.01).CONCLUSIONSSign‐out composition changes over time, and is associated with workload. Future interventions to improve quality should take these factors into consideration. Journal of Hospital Medicine 2015;10:534–536. © 2015 Society of Hospital Medicine.