Medical malpractice claims in Hepatology: Rates, Reasons, and Results

Background:

Clinicians are motivated to provide safe, high-quality care to patients with chronic liver disease. This includes the desire to avoid litigation. Data are limited regarding the actual sources of medicolegal risk in chronic liver disease.

Methods:

We conducted a review of a national liability insurer (Candello) with an additional granular analysis of our institution’s registry of liability claims. We included closed cases involving chronic liver disease-related encounters between 2012 and 2021. We determined rates of legal claims from a denominator of unique patients with cirrhosis or transplant care seen over the study period.

Results:

Local database: We retrieved 39 claims of which 15 involved patients with non-cirrhotic chronic liver disease, 13 involved cirrhosis (0.06% incidence), and 11 involved patients who underwent transplantation (0.6% incidence). Most claims involved periprocedural complications. Others included adverse reactions to prophylactic plasma transfusion, medication-induced HE, and falls/fractures.

National database:

We found 94 claims related to liver disease out of 102,575 (0.09%) total claims. Overall, 56% involved diagnosis-related issues (failure/delay in ordering a diagnostic test, failure to appreciate and reconcile a symptom/sign or result, or the misinterpretation of a diagnostic study). Miscommunication between providers and between providers and patients was implicated in 22% of cases. Patient behavior-related factors (nonadherence with scheduled appointments, treatments, or diagnostic testing) factored in 20% of cases. Selection or the management of therapy played a role in 7% of cases. Very rarely were cases associated with technical skill (4%), house staff supervision (3%), or weekend/holiday care (1%). Fifty-one (55%) claims involved HCC.

Conclusion:

We provide the rates and reasons for medical malpractice claims in hepatology.

INTRODUCTION

Chronic Liver disease (CLD) is increasingly common, complex, and costly.1 Accurate diagnosis of the underlying etiology and timely diagnosis of cirrhosis complications, such as hepatic encephalopathy (HE), ascites, and hepatocellular carcinoma (HCC) are necessary to avoid substantial patient harm. Liver disease care can be complicated and risky. CLD increases the risk of even routine medical care. In addition, patients with liver disease often require invasive procedures. The only true cure for cirrhosis is transplantation, which depends on a limited pool of donor organs requiring challenging decisions regarding transplant candidacy. Each factor raises the stakes of hepatology practice and imbues care with widespread concerns of legal liability. Data are limited, however, regarding liability and its sources.

Gastroenterology practice (which includes hepatology) is associated with the sixth highest per-physician rate of annual malpractice claims but a below-average rate of damages awarded per closed claim.2 In 2014, gastroenterology was associated with 12.1 claims per 1000 physician years.3 However, little is known about hepatology-specific claims. In an audit of all 85 claims against gastroenterologists from England and Wales, 1987-1996, only 1 was related to liver disease.4 From 1985 to 2005, 500 claims were made regarding procedural complications in gastroenterology, 50 of which were related to liver disease-related procedures (eg, biopsy).5 As of 2019, 1562 judicial opinions had been published regarding liver disease-related lawsuits. These focused on the denial of insurance based on alcohol abstinence, denial of waitlisting or hepatitis C therapy for prisoners, and rare claims of age or religious discrimination.6 Unfortunately, most prior reports lack the granular details required to inform clinicians about sources of liability and the practice improvements needed to avoid it. Without this granularity, lessons from these experiences are often lost rather than leveraged to improve quality and safety.

Herein, we review 2 sources to define the hepatology liability landscape and thus, the opportunities for care improvement: our institution’s experience and all cases from a national legal liability insurer.

METHODS

Our study consists of 2 main data sources: institutional data and national data. The institutional data allowed a detailed assessment of patient and provider factors and provided an estimate of the incidence of claims among our institutional cohort with CLD, cirrhosis, and transplant-related care. The national data was extensively coded for contributing factors and was assessed for conceptual drivers of liability risk. We evaluated all ‘closed’ cases from both sources, which refers to all cases with a conclusion, including settlement, dismissal, and withdrawal. This study was approved as exempt from review by the University of Michigan Medical School Institutional Review Board.

Local claims

We conducted a review of our institution’s registry of liability claims using a keyword search in the Executive Summary field. The terms are summarized in Supplementary Table 1, http://links.lww.com/HC9/A247. The search included loss dates between 2012 and 2021 with a status of ‘closed.’ Each case was rendered into a clinical summary from which the primary diagnosis and allegations were derived. All records retrieved were reviewed by 3 clinicians to determine if CLD and/or liver transplantation were involved in the case. Cases were determined to be related to cirrhosis if the patient had cirrhosis, liver transplant if it involved a patient who was undergoing posttransplant or a patient who was undergoing liver transplantation, and CLD if it involved liver diseases or diagnostics not involving patients with cirrhosis or liver transplantation. In addition, the responsible service(s), patient outcome, case status (closed or settled), and indemnity payment amounts were extracted. The incidence of claims was calculated for people with cirrhosis and liver transplantation. This was accomplished by searching the electronic medical records for patients with validated diagnosis codes for cirrhosis 7 and for all patients who had received or undergone a liver transplant. We could not determine a denominator for CLD as many (eg, NAFLD, alcohol-associated liver disease) have insensitive diagnostic codes.

National claims

We rendered all CLDs and cirrhosis complications into validated diagnostic codes7 to pull closed cases attributed to those diagnostic codes from Candello’s Comparative Benchmarking System. (Supplementary Table 1, http://links.lww.com/HC9/A247) Candello is a division of the Risk Management Foundation and the Controlled Risk Insurance Company. The Candello database represents ~one-third of all US malpractice claims from medical centers throughout the country, representing roughly 550 hospitals and health systems. Claims are analyzed by Candello-employed clinical experts, and major contributing factors to each malpractice claim are assigned. In this study, all claims with the loss year of 2012 through 2021 involving patients with final diagnosis codes relating to liver disease were reviewed. Basic demographic data were collected for each case, along with the responsible service, procedure, injury severity, indemnity payment amounts, and contributing factors. The coding of contributing factors is determined based on a multitiered taxonomy structure that is designed to capture the clinical and legal variables of each malpractice claim by utilizing medical records and legal claim files. The contributing factors are designed to identify opportunities for improvement or medical errors that led to the specific patient's injury or death. Case severity is defined using the National Association of Insurance Commissioners severity scale: low (0–2), medium (3–5), and high (6–9). Low scores suggest a legal issue alone; high scores are consistent with patient death.

RESULTS

Local claims

Our search returned 39 claims, 15 for patients with non-cirrhotic CLD, 13 with cirrhosis, and 11 with transplant-related conditions (Table 1). For patients with non-cirrhotic liver disease, the vast majority had cancer involving the liver, with claims related to surgical or periprocedural complications (eg, bleeding from a biopsy). There were 2 cases of nosocomial transmission of viral hepatitis during procedures and 1 patient who received a false diagnosis of liver disease due to lab error. Severe injury due to perioperative/periprocedural complications and DILI resulted in the settled claims, with mean damages of $304,000 (range $9,000–$1,100,000). Both cases of DILI resulted from the claim of inadequate monitoring after the drug initiation: 1 for isoniazid and 1 for disulfiram.

TABLE 1.

Institutional closed claims 2012-2021

Clinical context	N	Settled	Damages (mean)	Claims	Harm
Non-cirrhotic chronic liver disease	15	8	304,000	Nosocomial transmission of viral hepatitis (2), DILI (2), cancer therapy complication (3), postoperative complication (3), procedural complication (1), failure to treat/prevent thrombosis (2), false positive labs indicating liver disease (1), failure to manage cardiomyopathy leading to ischemic liver injury (1)	Death (6), viral hepatitis (2), pain and suffering (1), organ damage (3), no permanent damage (2), disease progression (1)
Cirrhosis	13	5	385,000	Postprocedural bleeding (3), postprocedural infection (2), failure to list for transplant (1), postoperative complication (2), opioid-induced HE (1), Benzodiazepine-induced HE and fall, contrast-induced AKI (1), prophylactic plasma transfusion complication (1)	Death (8), organ damage (1), extended hospitalization (1)
Liver transplantation	11	4	485,000	perioperative morbidity (4), intraoperative complication (3), in-hospital fall (2), biopsy complication (2)	Death (5), brain injury (2), reoperation (1), prolonged hospitalization (1), organ damage (1), none (1)

Notes: The settled cases refer to cases that conclude with damages paid to the plaintiff. Cases that have not been settled were either dismissed or withdrawn.

Liver transplant-related care resulted in 11 claims, including 4 settled claims, with mean damages of $485,000 (range $12,000-$1,130,000). The claims occurred among 1751 transplant patients cared for over the study period (0.6% incidence). All but 2 involved perioperative/periprocedural bleeding, infection, air embolus, and bile leak complications. There were 2 claims for in-hospital falls in the postoperative period.

Patients with cirrhosis accounted for 13 claims out of 22,230 unique patients seen during the study period (0.06% incidence). Five cases were settled with mean damages of $385,000 (range $12,000–$850,000). Most claims involved periprocedural complications like infection and bleeding. The bleeding complications arose from vascular injury (eg, jugular vein laceration during dialysis catheter placement, epigastric artery laceration during paracentesis, intercostal artery laceration during thoracentesis, and umbilical vein laceration during radiofrequency ablation). One patient experienced a reaction to a prophylactic plasma transfusion. There were 2 claims related to opioid-induced or benzodiazepine-induced HE resulting in falls or other complications that precluded transplantation. Failure to list for liver transplant in a timely fashion was claimed for 1 deteriorating patient, but the plaintiff's attorney closed the case, electing not to pursue it. Two patients experienced complications attributed to iodinated contrast agents: acute kidney injury and burns due to skin exposure. Finally, 1 patient developed respiratory arrest during nasogastric tube placement.

National claims

Overall, 94 closed claims relating to liver disease were retrieved from the Candello database out of 102,575 claims (0.09%) (Table 2). The severity of the injury in the claims was classified as “high” in 77% of claims, and 32% were settled with a median indemnity of $470,606. All liver diseases were searched, but not all were involved in claims. Fifty-two (55%) cases pertained to HCC, 12 (13%) related to hepatitis C (both acute and chronic), and other liver conditions accounted for a small number of cases.

TABLE 2.

National closed claims 2012-2021

	Closed Cases	Severity	Settled	Gross indemnity	Examples
	Total	High/Medium/Low		Median (US dollars)
Acute hepatitis C	6	1/4/1	3	42,500	Failure to assess for chronicity
Chronic hepatitis C	5	2/1/2	2	303,106	Failure to diagnose before decompensation
Hepatocellular Carcinoma (HCC)	52	48/4/0	18	650,000	Several cases of failure to screen with imaging, including failure to identify cirrhosis; 1 case of failure to communicate positive radiology findings.
Esophageal varices	4	3/1/0	0	N/A	Failure to diagnose varices endoscopically
Spontaneous bacterial peritonitis	3	2/1/0	0	N/A	Failure to interpret cell count studies as diagnosis of SBP
Alcohol-associated hepatitis	1	1/0/0	0	N/A	N/A
Alcohol-associated cirrhosis	6	3/3/0	0	N/A	Inappropriate use of antibiotics leading to fatal DILI
Autoimmune hepatitis	1	1/0/0	1	250,000	Failure to diagnose and treat seronegative AIH before jaundice
Nonalcoholic Fatty Liver Disease (NAFLD)	1	1/0/0	0	N/A	Fall and fracture at home due to untreated HE.
Unspecified chronic liver disease	1	1/0/0	0	N/A	Liver failure after TIPS
Hepatic Encephalopathy (HE)	3	2/1/0	0	N/A	Failure to prevent cerebral herniation in ALF.
Congenital liver disease	8	5/2/1	6	371,751	N/A
Ascites	3	2/0/1	0	N/A	Hemorrhage on uninterrupted systemic anticoagulation
Total	94	72/17/5	30	470,606	—

Notes: The settled cases refer to cases that conclude with damages paid to the plaintiff. Cases that have not been settled were either dismissed or withdrawn.

Abbreviations: AIH, autoimmune hepatitis; ALF, acute liver failure; DILI, drug-induced liver injury; SBP, spontaneous bacterial peritonitis; TIPS, transjugular intrahepatic portosystemic shunt.

Each case was assessed for its contributing factors. The most common was diagnosis (56%), driven by a failure/delay in ordering a diagnostic test, failure to appreciate and reconcile a symptom, sign, or result, and the misinterpretation or delayed interpretation of a diagnostic study. Miscommunication between providers and between providers and patients was involved in 22% of cases. Patient behavior-related factors (nonadherence with scheduled appointments, treatments, or diagnostic testing) were involved in 20% of cases. Selection or management of therapy played a role in 7% of cases. Very rarely were issues identified relating to technical skill (4%), house staff supervision (3%), or weekend/holiday care (1%).

Among the numerous (52, 55%) HCC cases, the most common reasons for liability—in 37 (71%) of cases—included the failure to screen candidates and to identify patients at risk for HCC (ie, failure to diagnose cirrhosis in a patient with known CLD). In addition, 40% of HCC claims involved failures of communication, for instance, the failure to read the medical record, failure to communicate radiology findings to the responsible clinician, and ineffective communication during transitions of care.

For the 6 (6%) cases relating to acute hepatitis C, a common thread was the failure to assess for chronicity resulting in CLD, while the 5 (5%) involving chronic hepatitis C resulted from a failure to diagnose (and treat) before decompensation. Procedure and treatment-related complications played a role in 13 (25%) of claims.

Other claims involved the quality of diagnosis, such as he failure to diagnose varices endoscopically, identify patients with spontaneous bacterial peritonitis based on cell counts, and initiate appropriate management in a patient with seronegative autoimmune hepatitis whose delayed diagnosis resulted in liver failure. Periprocedural complications were common, such as a case involving liver failure after transjugular portosystemic shunt and hemorrhage after paracentesis in a patient whose direct-acting anticoagulant was not temporarily held. The only case attributed to NAFLD was a case of undiagnosed HE that resulted in premature hospital discharge, a fall, and a fracture. HE resulted in another claim where cerebral herniation occurred in a patient with acute liver failure. Finally, DILI in a case where administered antibiotics were found to be inappropriate resulted in a claim involving a patient with alcohol-associated liver disease.

DISCUSSION

Though the desire to avoid litigation is a strong motivator for clinical decision-making, there is limited empirical data on which to ground this concern. Furthermore, this desire may detract from the broader aim of quality improvement and result in time spent away from the patient care and research activities for which most clinicians entered the profession. Data and guidance are needed. Our comprehensive review of medicolegal claims in hepatology practice advances the aims of quality care with 3 major findings. First, we detail the hitherto poorly described sources of legal claims in hepatology. Second, we define both the per-patient risk of legal claims in hepatology and the proportion of legal claims relating to liver care. Third, we provide several concrete areas for practice improvement informed by the risk of liability that could result in improved patient safety.

What leads to legal claims in hepatology and why?

We summarize our findings in Figure 1. The national data suggests that care discrepant from guidelines and the failure to recognize and respond in a timely fashion to liver-related conditions is the key driver of liability. Most claims involve liver cancer and a failure either to follow screening guidelines for at-risk patients or to correctly identify who should be screened. Other cases involve a failure to diagnose and treat conditions when patients present, allowing an undifferentiated condition to worsen (eg, hepatitis C, autoimmune hepatitis, spontaneous bacterial peritonitis, and HE). Taken together, these gaps in care can be addressed with quality improvement that aims to identify patients at risk for cirrhosis complications and link them to effective interventions. Tools such as best practice advisories can alert providers to the role of hepatitis C screening or liver cancer screening and may support clinicians (and quality improvement efforts) accordingly.8-10 Measures such as noninvasive testing like FIB-4 to trigger outpatient referral can indicate to providers when hepatology consultation may be beneficial. For hospitalized patients, pathways to standardize care and indicate the role of inpatient hepatology consultation for conditions such as HE or symptomatic ascites could improve outcomes. Similarly, alerts that flag the therapies that may be more toxic in patients with cirrhosis (ie, benzodiazepines or opioids) may provide just-in-time reminders or nudges to prescribers as a risk mitigation strategy. While the contributing factors and conceptual drivers of liability risk, including communication and delayed recognition, may not differ between claims observed from other medical conditions, the systematic responses required to address vulnerabilities may be specific to the care needed by patients with CLD.

FIGURE 1.

Sources of Medical Malpractice in Hepatology. A conceptual model synthesizes the structural, patient, and provider-related factors that are associated with liability claims in hepatology. HCC = hepatocellular carcinoma, INR =international normalized ratio, SBP = spontaneous bacterial peritonitis

In contrast, our local data reminds us that the specific sources of legal claims facing any given institution are variable, highlighting both strengths and gaps in care process. For example, falls relating to HE figures in both local and national data possibly represent widespread pitfalls in the management of HE. Conversely and notably, HCC (particularly the timely diagnosis of HCC) comprises most claims nationally, but Michigan lacks any claims relating to HCC screening. There are multiple potential explanations. First, care quality can be highly variable. Some centers may have more robust systems in place for care delivery than others. For example, the University of Michigan has had a prospective disease management software before the index date for this study that reminds clinicians when at-risk patients are due for HCC screening.11 This approach has been modernized into a system-wide dashboard with provider alerts.8

Second, the similarities between the local and national databases may explain which types of events are more likely to result in legal action regardless of the culture. Both datasets appear to suggest that the severity of the patient outcome influences the risk of a lawsuit. In the Michigan experience, most claims involved a patient's death. In the national dataset, 77% of claims involved a high-severity injury (inclusive of death). For this reason, outcome-driven claims like procedural complications and rare adverse events such as DILI can be found in both datasets. Indeed, prior studies have found that as many as 35% of claims do not relate to medical errors.12 In many cases, a proactive process of disclosure might be able to mitigate the risk of claims, but high-severity injuries may result in claims regardless.

Instructive cases

Our study highlights potentially overlooked sources of risk. First, we found multiple cases of falls because HE or after transplant. In the short term, fall risk can be reduced through psychoactive medication stewardship, recognition of and timely therapy for HE, and deliberate but cautious patient mobilization.13-15 In addition, a proactive approach to addressing malnutrition, frailty, and sarcopenia may reduce the risk of falls before and after liver transplantation.16 Second, the case of a transfusion reaction with harm is instructive. While multiple professional organizations recommend avoiding prophylactic blood product transfusions before low-risk procedures, plasma and platelets are frequently given to patients with cirrhosis undergoing paracentesis or upper endoscopy.17 This offers no benefits but exposes patients to risk. Understanding and reframing the liability risk related to blood product transfusion may help align clinical practices with guideline recommendations. Third, the risk of liability due to DILI highlights the importance of antibiotic stewardship and close monitoring of patients receiving potentially hepatotoxic therapies.

The rate of liver-related claims

In a national medicolegal liability claims database, CLD resulted in 94 cases or 0.09% (9 per 10,000) of the total volume of liability claims. At Michigan, 0.06% (6 per 10,000) of patients with cirrhosis were involved with a legal claim. The rate was 0.6% (60 per 10,000) among all patients who underwent liver transplantation. Benchmarks for rates of medical professional liability claims are lacking. Our data helps address this knowledge gap. First, rates of claims may not be related to the prevalence of the disease. Cirrhosis may impact 0.3–1% of the population and CLD at least 10%;18 even without adjusting for the severity of the disease and health care utilization, this suggests that the rate of CLD-related claims is lower than its prevalence. Second, when comparing with rates from other publications, it is important to consider the timeframe used for rate determination. While our incidence rates pertain to any claim per-patient seen over a 10-year period, prior studies have evaluated claims per encounter over much narrower cross-sectional periods. Indeed, 13 liability claims resulted from every 10,000 hospitalizations in a 1984 study from New York state 19 and 12 per 10,000 hospitalized patients in Utah and Colorado in 1992.20 Risk-management events (not necessarily legal claims) result from 37 per 10,000 primary care clinic visits.21 Patients with cirrhosis require frequent procedures and incur elevated risks of complications, which would intuitively increase the risk of claims. However, there may be reasons why care for a patient with CLD could result in lower-than-expected rates of liability claims, given the frequent social isolation and socioeconomic factors reducing access to legal support. This may explain why we found that claims are 10-fold more likely among patients undergoing a transplant. Patients undergoing liver transplantation are preselected based on social support and the ability to navigate the medical system while also incurring elevated risks due to the complexity of liver transplantation.

Preventative measures

The optimal methods to reduce the risk of liability claims are unclear. In addition to the condition-focused suggestions above, there are multiple systematic approaches that could be helpful. First, many claims result from complications of care or missed diagnoses. All interventions have risks, but each complication, when viewed closely, often had missed opportunities to course correct. A failure-to-rescue is an event where there is a failure to recognize or delay in responding to a patient experiencing a complication. Developing system-wide approaches to detect deterioration (eg, sepsis alerts for hypotension or other signs) and respond to clinical changes (rapid response teams) may result in timely care for a patient prior to an adverse event.

Second, centers may approach medical errors with different strategies explaining some of the variances in which cases result in claims. For example, Michigan developed in 2001 a model wherein the University of Michigan Health System performs active surveillance for medical errors, fully discloses found errors to patients, and offers compensation when it is at fault. This program, also referred to as the ‘Michigan Model,’ decreased legal claims, lawsuits per month, time to claim resolution, and costs of liability management.22 A culture of early reporting, comprehensive care review, honesty, transparency, and a focus on lessons learned favorably impacts outcomes for patients and providers alike. This model allows for timely and objective care reviews, opens necessary dialogue with patients/families to prevent conjecture/the need to hire an attorney, shares lessons learned along with changes made to prevent a recurrence, and promotes healing for all involved. The implementation of this model was associated with reduced lawsuits (2.13–0.75 per 100,000 patient encounters) and reduced costs for total liability (RR, 0.41 [CI, 0.26–0.66]), patient compensation (RR, 0.41 [CI, 0.26–0.67]). Though future studies are required to determine the sources of heterogeneity in claim type and burden, as well as the impact on hepatology claims, the way an institution handles medical errors may have an impact on the rate and type of liability claims that are opened.

Contextual Factors

These data must be interpreted in the context of the study design. First, regional variation in care delivery and quality may mean that the landscape of liability claims may differ in regions not included in this study. Second, the lower granularity of data available in the national claims data allows for fewer case-specific lessons regarding the sources of liability. Finally, it is important to observe that claim data, including claim frequency and settlement values, are significantly influenced by jurisdiction, the existence of damage caps and other laws that vary from state to state, and insurance coverage. We lack data on the cities/states involved in the national database.

CONCLUSION

This study has provided the rates and reasons for medical malpractice claims in hepatology. Liability claims in hepatology indicate several areas for proactive attention to improve care quality. These include identification of patients with cirrhosis, linkage to care for liver cancer screening, safe prescribing of potentially hepatotoxic medications, and fall prevention for patients with cirrhosis and in the peri-transplant period.