Reducing chest X-rays, telemetry

Summaries from ACP Hospitalist Weekly.

Intervention reduced routine chest X-rays in cardiovascular ICU, but not in medical ICU

An intervention that used clinician education, peer champions, and data feedback to reduce ordering of daily chest X-rays in two adult ICUs was associated with fewer orders of chest X-rays in the cardiovascular ICU, but not in the medical ICU, a single-center study found.

Image by Getty Images
Image by Getty Images

Researchers conducted the prospective, nonrandomized study from October 2015 to June 2016 at a U.S. academic medical center. The initiative included staff of the medical and cardiovascular ICUs (physicians, surgeons, nurse practitioners, fellows, residents, medical students, and X-ray technologists) and involved didactic clinician education and weekly emails with chest X-ray ordering rates. Peer champions coordinated data feedback and advocated for the intervention among colleagues. In the medical ICU, three internal medicine residents volunteered as peer champions; in the cardiovascular ICU, two acute care nurse practitioners served in this role.

Researchers analyzed chest X-ray ordering rates during the intervention period compared to control data from a historical baseline period (Oct. 1, 2014, to Sept. 15, 2015), as well as identified facilitators and barriers of deimplementation in qualitative analyses. Results were published in the February Journal of Hospital Medicine.

In the cardiovascular ICU, the median baseline chest X-ray ordering rate was 1.16 chest X-rays per patient per day (interquartile range [IQR], 1.06 to 1.28). During the intervention period, the rate decreased to 1.07 (IQR, 0.94 to 1.21; P<0.001). A time-series analysis suggested a small but statistically significant drop in ordering rates during the intervention period (P<0.001), followed by a slight, nonsignificant upward trend.

In the medical ICU, the median baseline chest X-ray ordering rate was 0.60 chest X-ray per patient per day (IQR, 0.48 to 0.73) among the physician care team and 0.39 (IQR, 0.21 to 0.57) among the acute care nurse practitioner team. Both rates stayed about the same during the intervention period. A time-series analysis suggested a small but significant downward trend in ordering rates during the baseline period (P=0.011 for the physician team; P=0.022 for the acute care nurse practitioner team), and trends during the intervention period were not significantly different.

The main facilitators of deimplementation were peer champion discussions about chest X-ray ordering during rounds and weekly data feedback, particularly if accompanied by in-person follow-up. Barriers included the need to establish goal chest X-ray ordering rates, insufficient intervention visibility, and waning investment among residents in the medical ICU due to frequent rotation and competing priorities.

The estimated cost of the initiative was $1,600 ($1,000 for curriculum development and teaching time, $300 for educational materials, and $300 for chest X-ray tracking dashboard development), and the annual cost to continue the program was estimated to be $1,000 per year. An informal, conservative estimate of savings from the program showed a net savings of $7,280 to $17,000 per year in the ICU.

“Although these amounts are modest, providing trainees with experiential learning opportunities in high-value care is valuable in its own right, meets curricular goals, may result in spill-over effects to other diagnostic and therapeutic decisions, and may influence long-term practice patterns,” the study authors wrote.

Limitations of the study include its single-center design and lack of a concurrent control group. In addition, the outcome was all chest X-rays ordered, regardless of necessity, and researchers did not track clinical indications for chest X-rays.

Bundled payments for hip, knee replacement reduced costs with no change in complications

The implementation of mandatory bundled payments for joint replacement was associated with reduced spending and no change in complication rates, a recent study found.

The study analyzed the effects of Medicare's Comprehensive Care for Joint Replacement (CJR) program, which, starting in 2016, randomly assigned hospitals in some areas to receive bundled payments for hip or knee replacement. Using Medicare claims from 2015 through 2017, the authors compared 280,161 hip- or knee-replacement procedures at 803 hospitals in 75 metropolitan statistical areas that were assigned bundled payments to 377,278 procedures at 962 hospitals in 121 control areas that didn't have bundled payments for joint replacement. Results were published by the New England Journal of Medicine on Jan. 2 and appeared in the Jan. 17 issue.

The study found greater decreases in institutional spending per joint-replacement episode in the treatment areas than in control areas; the between-group difference was −$812, or a −3.1% differential decrease (P<0.001). This difference was driven largely by a 5.9% decrease in patients discharged to post-acute care facilities. There were no significant differences in the composite rate of complications (P=0.67) or the percentage of joint-replacement procedures performed in high-risk patients (P=0.81).

The mandatory nature of the CJR model made it an unusual and controversial experiment in payment reform, the study authors noted. The Trump administration changed it to a partly voluntary program in March 2018, but the data used in this study help answer the question of whether benefits seen from previous bundled payment programs were attributable to hospitals volunteering to participate. “Our findings suggest that the changes observed in voluntary programs may be echoed in mandatory programs,” the authors wrote. They also found that spending was decreasing over the studied 18 months, suggesting that greater cost reductions might eventually be seen from the program.

The authors said that it is not surprising that the savings mostly came from reductions in the use of post-acute care services, because this is a large and highly variable component of the cost of joint replacements, and “hospitals may have successfully identified patients who are at the margin of needing post-acute care services who could instead be safely discharged home with home health services.” They noted that the study did not include patient-centered outcome measures, including functional status, pain, and satisfaction. Other limitations of the study include that its results may not be generalizable to bundled payment programs other than hip- or knee-replacement procedures.

AHA program appears to have improved STEMI care over time

Quality of care for ST-segment elevation myocardial infarction (STEMI) appeared to improve over time in hospitals participating in the American Heart Association's (AHA) Mission: Lifeline Program.

The AHA's Mission: Lifeline Program aims to improve guideline-recommended STEMI care by developing efficient, organized, and coordinated health care systems in the community. Hospitals volunteer to participate in Mission: Lifeline. All patients with STEMI who were admitted to Mission: Lifeline hospitals from Jan. 1, 2008, to Dec. 31, 2012, were entered into the National Cardiovascular Data Registry Acute Coronary Treatment and Intervention Outcomes Network Registry—Get With the Guidelines Registry, which provides data collection and evaluation for Mission: Lifeline. The current study examined prehospital and in-hospital care and outcomes from 2008 through 2012 in patients with STEMI. In-hospital adjusted mortality rates were calculated with and without cardiac arrest as a reason for delay in primary percutaneous coronary intervention (PCI). The study results were published Dec. 31, 2018, by the Journal of the American Heart Association and appeared in the Jan. 8 issue.

Overall, 147,466 patients at 485 hospitals were included in the study. Thirty percent of patients were women, and the median age was 60 years. Over the study period, a decrease was noted in the proportion of eligible patients who did not receive reperfusion (6.2% vs. 3.3%) and fibrinolytic therapy (13.4% vs. 7.0%). The median time from symptom onset to first medical contact remained at approximately 50 minutes during the study period, while use of prehospital electrocardiography increased (45% vs. 71%). Improvements were seen over the study period in median first medical contact-to-device time for emergency medical systems transport to PCI-capable hospitals, median first door-to-device time for transfers for primary PCI, and median door-in/door-out time at non-PCI-capable hospitals (P<0.001 over the study period for all comparisons). Overall in-hospital mortality increased over the study period (5.7% vs. 6.3%). At three years and five years, adjusted mortality rates excluding patients with known cardiac arrest decreased by 14% and 25%, respectively (P<0.001).

The researchers noted that the generalizability of their findings may be limited because hospital participation in Mission: Lifeline is voluntary and that changes over time may reflect changes in patient population as well as in patient care, among other limitations. They concluded that quality of care for STEMI improved over time in Mission: Lifeline hospitals and that while in-hospital mortality improved in patients without cardiac arrest, it did not appear to improve overall as numbers of high-risk patients increased. “This highlights the need for improving system-level care for out-of-hospital cardiac arrest, an ongoing objective of Mission: Lifeline,” the researchers wrote.

Electronic health record alert reduced unnecessary telemetry outside the ICU at one hospital

Adding an alert in the electronic health record (EHR) significantly reduced unnecessary inpatient telemetry in a recent single-center study.

The cluster randomized trial included 12 hospital medicine (eight housestaff and four hospitalist) teams at the University of California, San Francisco, Medical Center. Between November 2016 and May 2017, half the teams were randomized to receive an EHR alert when a patient's telemetry order exceeded the duration recommended for the indication. The alert displayed only during daytime hours and for patients not in the ICU. Results were published by JAMA Internal Medicine on Dec. 10, 2018, and appeared in the January issue.

The study included 1,066 hospitalizations of 1,021 patients. The intervention arm had 499 hospitalizations (mean age, 64.5 years; 45% women) and the control arm, which provided usual care, had 567 hospitalizations (mean age, 63.8 years; 46% women). Compared to controls, the patients whose teams received alerts had a significantly shorter duration of telemetry monitoring (−8.7 hours per hospitalization; 95% CI, −14.1 to −3.5 hours; P=0.001) with no significant change in rapid-response calls or medical emergency events. In response to the 200 alerts that occurred, physicians discontinued telemetry 62% of the time.

Previous research has targeted the same goal using other and multicomponent interventions, and this study extends those efforts, according to the authors. “First, we demonstrate that an EHR alert targeted to the right clinicians (primary team physicians), at the right time (daylight hours, while the physician is writing orders), with the right clinical context (telemetry exceeds recommended duration) can succeed without an associated educational or audit and feedback campaign,” they said. The reduction in telemetry from the EHR alert was smaller than that from multicomponent projects, but it is also a less costly and more scalable intervention, the authors noted.

The generalizability of the results is uncertain because this was a single-center study. The intervention was also more permissive of telemetry than national practice standards, allowing indications such as acute pulmonary embolism and severe sepsis, the authors said. However, they noted that the hospital's baseline mean telemetry hours were also lower than the baselines in other similar studies, which may have limited the effect size.