Cirrhosis is a prevalent and expensive condition, affecting approximately 2.5 million individuals at a cost of over $4 billion annually in the United States.1–3 Cirrhosis is the second leading cause of digestive disease–related mortality, preceded only by colorectal cancer.3 The human and economic burden of cirrhosis is expected to increase substantially in the next 10 years as a result of an aging chronic hepatitis C population4 and the possible consequences of fatty liver disease.
Appropriate medical care for cirrhosis can delay complications, improve quality of life, and possibly extend survival.5–8 Experts commissioned by professional societies have published comprehensive evidence-based guidelines to help clinicians better manage these patients.9, 10 Existing data are sparse in the extent to which patients with cirrhosis receive guideline-recommended care, but few important studies indicate significant shortfalls.5, 11 In a cohort of 774 patients with cirrhosis and ascites seen at three Veterans Affairs Medical Centers between 2000 and 2007, we recently found that nearly all patients with documented spontaneous bacterial peritonitis (SBP) received antibiotics for treatment of SBP.12 However, only 30% of these patients received recommended antibiotics for secondary prophylaxis after their discharge from the hospital. In general, care targeted at diagnosis of acute events, especially with regard to hospitalized patients (e.g., paracentesis, ascitic fluid analysis) and treatment (e.g., antibiotics for documented SBP, diuretics), is more likely to meet guideline-recommended standards than preventive and more elective outpatient care (e.g., primary and secondary antibiotic prophylaxis for SBP), despite extensive data that preventive care improves outcomes in patients with cirrhosis.
The first step toward improving the quality of cirrhosis care is to try to measure it. Using a modified Delphi method, we developed a set of explicit quality indicators for patients with cirrhosis.13 These indicators can either be applied retrospectively (using preexisting data) or measured in real-time (at the time of clinical encounter). Regardless of the approach, the use of electronic health records (EHR)-based tools is likely to be important for efficient capture and reporting of quality measures. However, our studies show that the use of automated data that can be readily generated from EHR (such as diagnosis and procedure codes), while reasonably valid for some quality measures in cirrhosis, is fraught with errors for other indicators. For example, there are many reasons why patients do not receive recommended care, some of which are justifiable exceptions (e.g., presence of comorbidities, contraindications, and patient refusal).12 We found that for some of the cirrhosis quality indicators, preexisting data did not capture these exceptions to allow for accurate automated measurement.12 These results suggest that quality indicator measurement systems (such as EHR vendors) will need to incorporate ways of documenting justifiable exclusions. Clearly, more work is required to standardize and validate the methods of collecting cirrhosis quality indicators.
Quality measurement and reporting alone may not achieve higher quality care, and therefore efforts to improve care will need to target actionable drivers of timely cirrhosis care. Domains that can serve as potential targets for intervention at a practice level include providing better access to specialist care.12 Patient subgroups that may be important targets for future interventions include those with medical comorbidities.12 Our data also suggest that focused efforts on preventive care in ascites may improve quality of care delivered to patients with cirrhosis and ascites.12
Implementation of solutions and interventions will depend on the underlying problem as well as the structure and context of the clinical setting and practice needs (Table 1). Simpler interventions such as standardization (e.g., implementation of electronic order sets or care pathways) may be effective enough for situations that do not depend on the participation of many parties (such as ordering antibiotics in the setting of gastrointestinal bleeding and hepatocellular carcinoma screening). Designing a more expensive form of coordination may be necessary if the goal is to improve complex endpoints (such as hospital admission rates, morbidity, and mortality) that are dependent on many individuals, including patients, clinicians, caregivers, case managers, and so forth. We may also need to move toward new models of care in cirrhosis.14 These models include: (1) the application of elements of a chronic care model, in which active disease management continues in between clinic visits using home-based interventions such as home visits, scheduled telephone calls, and remote home monitoring systems; (2) multidisciplinary care, including gastroenterologists/hepatologists, primary care physicians, radiologists, infectious disease physicians and/or psychiatrists (who can either be colocated in a single clinic or actively engaged in cross-discipline collaboration via regular meetings [similar to tumor or transplant evaluation boards]); and (3) improving access to specialty care, particularly for patients with cirrhosis in underserved areas, via electronic consults and telemedicine. Although these models remain untested in patients with cirrhosis, their success in other areas of medicine suggest that they may improve outcomes in cirrhosis.
|Administration of antibiotics in all patients with cirrhosis and gastrointestinal bleeding||Standardized electronic order sets|
|Screening for hepatocellular cancer||Established care pathways|
|Reducing preventable hospital readmissions; improving morbidity and mortality||Improving access to specialty care|
|Multidisciplinary care, including gastroenterologists/hepatologists, primary care physicians, radiologists, and psychiatrists|
|Case management, including home-based interventions such as home visits, scheduled telephone calls, and remote home monitoring systems|
In our setting (an academically affiliated Veterans Administration Medical Center with a mature EHR), we are increasingly relying on clinical templates that incorporate clinical guidelines and small registries to monitor timely and consistent delivery of recommended preventive care (such as variceal screening, liver cancer screening, and vaccinations). We are expanding our clinical staff to include midlevel providers who, in addition to sharing the physicians' workload, are playing a major role in the coordination of patient care. We are also streamlining electronic communication between primary care providers and gastroenterologists through explicit follow-up recommendations for patients seen in the specialty clinic and detailed justification in cases where consults are discontinued. These are some examples of steps we are taking to improve quality of care delivered to patients with cirrhosis while waiting for more concrete evidence regarding the effectiveness of alternative care models in this population.