Care transitions, hypertension, and more

Summaries from ACP Hospitalist Weekly.

Intensive transition intervention failed to reduce readmissions of high-risk patients

A hospital transition program for high-risk patients did not significantly affect readmission rates, a real-world study found.

The nonblinded trial included 1,876 patients from two hospitals in Charlotte, N.C., who were identified as having at least a 20% risk for readmission. Usual care was provided to 941 control patients, while 935 were assigned to a Transition Services (TS) program that bridged inpatient, outpatient, and home settings, providing virtual and in-person access to a dedicated multidisciplinary team for 30 days. Patients were excluded if they lived more than 150 miles from the hospital; were admitted from or discharged to hospice, a skilled nursing facility, or jail; left the hospital against medical advice; had acute psychiatric illness or drug dependency; had an active diagnosis of cancer or sickle cell disease; or had recent or current access to this or another transition program.

In an intention-to-treat analysis, the 30-day readmission rate was 15.2% in the TS group versus 16.3% in the usual care group (P=0.52), and there were no significant differences in 60- and 90-day readmission rates or 30-day ED visit rates. However, because the study had a very high crossover to usual care (74.8%), the researchers also conducted a per-protocol analysis of the 236 patients who actually participated in TS versus the 941 who received usual care and found a lower 30-day readmissions among the former (10.6% vs. 16.3%, relative risk [RR], 0.65 [95% CI, 0.44 to 0.97]; P=0.03).

The intention-to-treat analysis did find significant effects in some prespecified subgroups, although the authors cautioned that the low participation rate meant that the results could be due to chance. Patients who were referred to TS and readmitted had a lower rate of ICU admission (15.5% vs. 26.8%; RR, 0.74 [95% CI, 0.59 to 0.93]; P=0.02). In addition, patients with a primary diagnosis of sepsis had lower 30-day readmissions if they were assigned to TS (7.9% vs. 16.3%; RR, 0.49 [95% CI, 0.24 to 0.97]; P=0.03). The results were published by the Journal of General Internal Medicine on Aug. 14, 2018.

The results of this first prospective, real-world study to evaluate a complex transition intervention's effect on readmissions in the U.S. may not be surprising, given lack of significant improvement in national readmission rates, the authors said. The results show that “efforts to reduce readmissions are challenging to implement in the real world where a system is accountable for an entire population, not merely patients who are engaged and who participate,” they said.

The authors noted that even offering patients the intervention was challenging in practice: One-third of the patients referred to TS were not contacted prior to discharge for reasons such as being away from their room, not feeling well enough to talk, or being discharged before the TS navigator arrived. “Health systems implementing bridging interventions may need to increase navigator staffing or expand navigator reach with virtual technology,” the authors said. In addition, health systems and researchers working to reduce 30-day readmissions should look at “novel approaches designed to address the reality of ubiquitous, low participation rates in real-world implementation,” they said.

Antihypertensive treatment intensified at discharge regardless of outpatient blood pressure history

About 14% of older adults with hypertension who were admitted for noncardiac conditions had antihypertensive treatment intensified at discharge, even though half of them had well-controlled blood pressure prior to hospitalization, a recent study found.

Using data from the U.S. Veterans Administration (VA) Health System, researchers assessed how often patients ages 65 years and older with hypertension were discharged with intensified antihypertensive treatment after admission for an unrelated medical condition. Treatment intensification was defined as receiving a new or higher-dose antihypertensive agent at discharge compared with medications used prior to admission. Results were published online on Sept. 12, 2018, by the BMJ.

Out of 14,915 older adults (median age, 76 years) hospitalized for common noncardiac conditions in 2011 to 2013, 65% had well-controlled outpatient blood pressure before admission. Of 2,074 patients discharged with intensified antihypertensive treatment, 52% had well-controlled blood pressure before admission.

Image by Getty Images
Image by Getty Images

After adjustment for potential confounders, higher inpatient blood pressure was strongly associated with antihypertensive regimen intensification at discharge. Among patients whose blood pressure was well controlled prior to admission, treatment was intensified at discharge in 8% (95% CI, 7% to 9%) of those without elevated inpatient blood pressure, 24% (95% CI, 21% to 26%) of those with moderately elevated inpatient blood pressure, and 40% (95% CI, 34% to 46%) of those with severely elevated inpatient blood pressure.

There were no differences in rates of intensification among patients least likely to benefit from tight blood pressure control (limited life expectancy, dementia, or metastatic malignancy) or among those most likely to benefit (history of myocardial infarction, cerebrovascular disease, or renal disease).

“Our findings indicate that decisions to discharge patients with intensified antihypertensive regimens are likely driven by inpatient blood pressure readings and not the overall context of older adults' health or long term disease control,” the study authors wrote.

Limitations of the study include the possibility that pharmacy records missed drugs obtained outside the VA system and the inability to assess the frequency of drug discontinuation, the authors noted. They added that antihypertensive treatment may be intensified for reasons other than hypertension and that the VA patient population is primarily male.

While this study did not assess whether intensified blood pressure control resulted in drug-related harm, data from other studies suggest higher rates of adverse drug events with more aggressive blood pressure control (e.g., hypotension, syncope), an accompanying editorial added.

“Overall, clinicians would be wise to adopt Sin City's famous tagline, ‘What happens in Vegas, stays in Vegas;’ often the safest approach to inpatient chronic disease management should be to let what happens in hospital stay in hospital,” the editorialists wrote.

Infections related to injection drug use may be underreported in ED, hospital settings

More than half of infections treated in hospitals that are related to injection drug use may be unrecorded in existing surveillance estimates, according to a new study.

The retrospective cohort study used data on inpatient and ED visits from the Healthcare Cost and Utilization Project for California (2005 to 2011), Florida (2005 to 2014), and New York (2006 to 2013). Researchers identified all patients ages 12 years and older who were diagnosed with four types of infection or infection syndrome possibly associated with injection drug use: bacteremia or sepsis, endocarditis, osteomyelitis or septic arthritis, and skin or soft-tissue infection. To estimate the incidence of infections related to injection drug use, they identified cases in which drug use was recorded at the time of infection and cases in which drug use was not recorded at the time of infection but was recorded within six months before or after the infection diagnosis. Results were published online on Sept. 11, 2018, by Clinical Infectious Diseases.

A total of 11,245,342 visits had a diagnosis for any of the four types of infection. About 92% (n=10,331,191) of these infections had no associated drug use diagnosis within six months, while 370,830 (about 3%) had a concurrent drug use diagnosis and 543,321 (about 5%) had drug use documented within six months of infection but not at diagnosis. After incorporation of drug use identified in the six-month window, the researchers' estimated yearly incidence of infections related to injection drug use increased between 105% and 218%.

Unrecorded drug use was more likely in patients on Medicare, admitted to EDs, and with homelessness or housing problems, as well as at hospitals with a lower incidence of visits related to injection drug use. In contrast, injection drug use was less likely to be unrecorded in patients with hepatitis C virus infection or whose primary payer was self-pay or no charge. Teenagers were more likely to have unrecorded injection drug use compared to adults between ages 18 and 34 years. The likelihood of injection drug use being unrecorded increased with age for age groups 35 to 44 years, 45 to 55 years, and greater than 55 years compared to patients ages 18 to 34 years.

A limitation of the study is the possibility that unrecorded infections related to injection drug use represent incorrect or underutilization of diagnostic codes for drug use, rather than accurately reflecting the medical record, the study authors noted. They added that they did not have access to medication or medical record data to help validate diagnostic codes.

The authors said future work should study downstream effects associated with unrecorded injection drug use versus drug use identified at the time of infection. “Patients whose drug abuse is identified earlier may begin treatment earlier, recover faster, or experience fewer future and recurrent infections. If such downstream effects are associated with identifying [injection drug use] at the point of infection, our findings suggest there may be many opportunities to reduce the morbidity and mortality attributable to [injection drug use],” they wrote.

Persistent cognitive impairment after ICU stay associated with modifiable risk factors

Patients who develop new cognitive impairment after an ICU stay commonly have certain risk factors, some of which may be modifiable, a recent study found.

The observational case-control study included 98,227 adult patients treated in Mayo Clinic ICUs between July 1, 2004, and Nov. 20, 2015. Using previously validated measures of dementia and cognitive impairment, the researchers determined which patients developed new and persistent cognitive impairment between three and 24 months after ICU discharge. Results were published by Critical Care Medicine online Sept. 14, 2018, and in the December 2018 issue.

Overall, 22% of the ICU patients were found to be cognitively impaired, but only 2.5% had new and persistent impairment after the index ICU admission. Compared with 2,401 age- and sex-matched controls who had ICU stays but no impairment, those who developed impairment had higher chronic illness burden (Charlson Comorbidity Index, 6.2 vs. 5.1; P<0.01) and were more likely to have multiple ICU stays (22% vs. 14%; P<0.01). After adjustment, new and persistent dysfunction was associated with certain events in the ICU: acute brain failure, severe hypotension, hypoxemia, hyperthermia, fluctuations in serum glucose level, and treatment with quinolones or vancomycin. Sepsis was initially associated with new cognitive impairment but the association did not persist after adjustment.

The study authors noted that the observed rate of new and persistent cognitive impairment was lower than that found in other research. “Previous studies that did not take into account baseline or included short-term cognitive dysfunction may have overestimated the frequency of persistent cognitive impairment; alternatively, lack of sufficient follow-up in patients who became disabled after the ICU admission may have led to an underestimation of the frequency of persistent cognitive impairment in our cohort,” they said.

The primary goal of this study, however, was to identify risk factors that might be modified during the ICU stay, the authors noted. Based on the results, these factors may include prolonged fever, hypoxemia, severe hypotension, acute brain failure, dysglycemia, and treatment with quinolones and vancomycin. Preventive strategies could be targeted at these factors if the associations found by this study are confirmed by additional research, the authors suggested. The generalizability of the study was limited by its use of a suburban, elderly U.S. population with high access to ICU services, they noted.