What guidelines?: attitudes and behaviors for prescribing prophylactic hepatic encephalopathy medications

INTRODUCTION Many, but not all, patients with decompensated cirrhosis require portosystemic encephalopathy (PSE) prophylaxis medications. Lactulose can treat PSE and prevent secondary PSE, which helps prevent mortality.1 The American Association for the Study of Liver Diseases (AASLD) in conjunction with Choosing Wisely, an initiative of the American Board of Internal Medicine, published guidelines in 2014.2–4 The second recommendation states: do not continue treatment for HE indefinitely after an initial episode with an identifiable precipitant. This includes infection or variceal bleeding and possibly malnutrition.5 Similarly, although AASLD Practice Guidelines recommends secondary prophylaxis after an episode of overt PSE (Grade I, A, 1),3 it also states that the prophylactic therapy may be discontinued under circumstances where precipitating factors have been well controlled (Grage III, C, 2). To this end, a quality improvement study sought to understand behavior surrounding the AASLD Choosing Wisely recommendations through a plan-do-study-act (PDSA) cycle. METHODS This study comprising a qualitative survey, retrospective chart review, and quality improvement initiative was reviewed in March 2022 by the institutional Quality Improvement Committee and deemed to meet requirements as quality improvement and deemed exempt from Institutional Review Board approval. A 7-question survey (Appendix, https://links.lww.com/XCL/A9) investigating awareness and agreement of Choosing Wisely guidelines was conducted among internal medicine residents and attendings, and gastrointestinal (GI) and liver fellows and attendings recruited through convenience samples at a major academic medical center with high liver transplant volume. A retrospective chart review from 2019 to 2021 was performed on admitted patients who were prescribed lactulose. Data were extracted on inpatient prescriptions of lactulose, rifaximin, or zinc. Polyethylene glycol 3350-electrolyte solution was excluded given the high frequency of use outside of cirrhosis. Demographic and clinical information was collected, including inciting events such as variceal bleeding or infection. Malnutrition data were collected but not included in the analysis as was not a modifiable variable in the time period. Following the Institute for Healthcare Improvement for quality improvement,6 a PDSA cycle was conducted with educational interventions from March to September 2022, including education during the survey and email reminders. RESULTS Survey results There were 85 survey respondents, including 27 internal medicine residents, 16 internal medicine and preliminary interns, 9 GI attendings, 4 GI fellows, 9 hepatology attendings, 2 hepatology fellows, and 18 hospitalists. Of the respondents, 43.5% reported beliefs that all patients with cirrhosis with prior PSE should be on indefinite lactulose. Almost half, 47%, had not heard of choosing wisely, and 17.6% reported being aware but uninformed of hepatology-specific guidelines. There were no differences in awareness between hepatologists and nonhepatologists. After education during the survey, 65.9% of respondents agreed with PSE guidelines, 29.4% were unsure, and the remainder disagreed. Of the 83 responses, 67.5% reported not following guidelines, but they would adjust; 9.6% did not follow the guidelines with no plan to change; and 22.9% already followed guidelines. More nonhepatologists reported that, after this survey, they would incorporate recommendations into their practice (53/74) than hepatologists (3/9), as seen in Figure 1 (p = 0.014).FIGURE 1: Percentage of hepatologist and nonhepatologist responses to the survey question regarding the adoption of Choosing Wisely guideline on PSE prophylaxis medication.Chart review Overall, 276 admissions included 97 (35%) for patients previously on lactulose, 74 (27%) who were started and continued on lactulose, and 81 (29%) who were started on lactulose but stopped before discharge, and the remainder were discontinued. Rifaximin was started in 119 patients, continued in 59 (50%), and 78 (66%) who were given rifaximin during admission but were not continued on discharge. Zinc was also started on 11 patients and continued for 5 (46%). Initial prescriptions for lactulose, not on admission medication reconciliation, were seen in 155 patients, and Table 1 demonstrates general information for this sample population. TABLE 1 - Demographic information for the 155 unique patients prescribed first-time lactulose between 2019 and 2021 N (%) Single admission (2017–2019) 146 (94) Median age (IQR) 64 (51, 70) Sex (male) 100 (65) Inciting event 126 (81) GI bleeding 52 (41) Infection 64 (51) Cirrhosis etiology Alcohol-associated liver disease 82 (53) NASH 23 (15) HCV 19 (12) HBV 11 (7) AIH 3 (2) Other (DILI, HCC, PSC, PBC, PVT, Budd Chiari, cryptogenic cirrhosis) 17 (11) Ethnicity Non-Hispanic 81 (52) Hispanic 74 (48) Race African American 19 (12) White 52 (34) Asian 12 (8) Unknown 11 (7) Other 61 (39) Language English 114 (74) Spanish 25 (16) Russian 6 (4) Cantonese 2 (1) Mandarin 2 (1) Bengali 1 (1) Other 5 (3) TIPS No 136 (88) Prior (including admission for revisions) 19 (12) Abbreviations: AIH, autoimmune hepatitis; GI, gastrointestinal; IQR, interquartile range; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis. Overall, 81% of patients prescribed initial lactulose had an inciting event. For patients with initial lactulose prescriptions, there were no statistically significant differences for inciting events on lactulose continuation (overall chi-square 0.789, p = 0.37; infection: 0.654, p = 0.419; and variceal bleed: 0.113, p = 0.736). There were no statistically significant differences for inciting events on whether or not lactulose was initiated and continued on discharge by ethnicity (p = 0.216), race (p = 0.612), language (p = 0.144), sex (p = 0.672), or cirrhosis etiology (p = 0.658). Of the 29 patients who were started on lactulose for HE in the absence of an inciting event, 16 were continued on lactulose on discharge. This included 17 patients (58.6%) who were staffed by hepatology attendings. Quality improvement intervention From March 15 to September 15, there were 86 unique patients admitted with 7 remaining admitted at the end of the study period. Of the 78 discharged patients, 47 were male, 55 spoke English as their primary language, and 23 identified as White. Most were on lactulose before admission (43/86), while 16 were started on lactulose inpatient and continued on it on discharge. Among these 16 patients newly on lactulose, 2 did not have documented PSE, 2 had GI bleeding, and 5 had an infection (Figure 2). There were 40 patients discharged on rifaximin, 15 of whom were not previously taking it. Four patients were on zinc previously, and 5 patients were discharged on it.FIGURE 2: Flow diagram of patients during the quality improvement period.DISCUSSION Despite the 2014 publication, this study demonstrates little awareness, agreement, or adoption of the PSE medication guidelines. The potential rationale for survey discordance includes lack of education and awareness but disagreement. Hepatologists appear less influenceable by decade-old guideline recommendations without updated, higher quality supportive evidence given the potential risks and severity of PSE. Although comforting that prescribing practices did not elucidate inequity, perhaps, clinical guidelines suggesting cessation are insensible to practice, and more needed are guidelines on lactulose initiation. The PDSA cycle identified that 16 patients continued on lactulose despite 2 not having evidence of PSE and only 5 with PSE with an inciting infection, which, according to guidelines, could have been stopped. Quality measures referencing PSE could include documentation of precipitating factors.7 Specific mechanisms could be forced into clinical workflow to improve documentation for future PDSA cycles.8 However, it is also plausible that there was no documentation due to guideline disagreement. Limitations include limited time for the intervention to improve guideline awareness though they have been published for nearly a decade. The survey is limited given the convenience sample at a solitary institution. Nevertheless, this center is a high volume of liver transplants and is potentially representative. The data collected do not consider if the encounter was for a secondary PSE event, which could provide a rationale for continued prophylactic therapy, or if lactulose was subsequently discontinued outpatient after discharge. This quality study does not seek to answer if prophylaxis for PSE should be stopped, but rather what the attitudes and behaviors are surrounding recommendation adherence. Results suggest that, despite established guidelines, there remains a lack of knowledge, adoption, and discussion of these recommendations. The guidelines were published nearly a decade ago to promote high-value care merit re-evaluation.