The Memphis, Tennessee, metropolitan area is experiencing an epidemic of hepatitis A. Memphis and Shelby County have a population of approximately 830 000 persons, 50% of whom are African American. For each year from 1989 to 1993, an average of 57 cases of hepatitis A were reported, but 210 cases were reported in 1994 and 1538 cases were reported in 1995 [2]. Neighboring counties in Mississippi and Arkansas noted similar increases [3]. This epidemic has not been associated with a common food source or a breakdown in public sanitation. Few recent urban epidemics of hepatitis A in industrialized countries have not been traced to a common source [4]. We present an analysis of the clinical features of patients hospitalized during this epidemic.

Methods

Top Methods Results Discussion Author & Article Info References

The computerized medical records of the 15 acute care hospitals (2 of which are pediatric hospitals) in Shelby County were queried for the discharge diagnosis of acute hepatitis A (International Classification of Diseases, Ninth Revision [ICD-9] codes 070.1 [acute hepatitis A without coma] and 070.0 [acute hepatitis A with coma]) for the study period of 1 January 1994 through 31 December 1995. The paper charts for the retrieved patient names were reviewed by five of the authors, who used a standardized form to record demographic data, epidemiologic risk factors, reasons for hospitalization, underlying medical conditions, and serious complications. A diagnosis of acute hepatitis A required positivity for HAV IgM antibody in conjunction with compatible symptoms and laboratory findings. We sought information on the following risk factors for transmission of HAV: exposure to day care, household contact, association with a restaurant, being a homosexual man, occupational exposure, hemophilia, active use of injection drugs, travel to areas in which hepatitis A is endemic, consumption of raw shellfish, and unknown. We also sought information on underlying medical conditions, including chronic liver disease, significant alcohol consumption (>50 g/d for >5 years), HIV infection, cardiovascular disease, underlying cancer, chronic immunosuppression, and diabetes.

Complications sought included acalculous cholecystitis, hemolysis, autoimmune hepatitis (antinuclear antibody titer > 1:160 or positive smooth-muscle antibody titer > 1:80 and compatible liver biopsy findings), pancreatitis, prolonged cholestasis (serum bilirubin level > 136 µmol/L [8 mg/dL] for >12 weeks), and relapsing hepatitis. If two or more unrelated complications were present, they were all recorded. If a complication was caused by an underlying condition (for example, acute renal failure in a patient with fulminant hepatitis), only the underlying complication was recorded. Any unanticipated complication that was encountered during chart review was recorded.

This study was approved by the institutional review board of the University of Tennessee. The Fisher exact test was used to analyze the data. Statistical significance was defined as a P value less than 0.05.

Results

Top Methods Results Discussion Author & Article Info References

We identified 313 patients and reviewed their charts. Fifty charts were excluded because patients lacked documented HAV IgM antibody. Four charts could not be retrieved. Three patients who were seen in two hospitals were not counted twice. The remaining 256 charts belonged to patients who met the criteria for hepatitis A. Demographic data for the study patients are shown in Table 1. Two hundred thirty-one patients were residents of Shelby County, and 25 were from surrounding communities.

The source of hepatitis A was identified in only 21% of patients (Table 1). The two restaurant-associated cases were from a large outbreak in Arkansas.

On admission, 89% of patients had prolonged nausea or vomiting, or both; 26% had coagulopathy (prothrombin time 3 seconds prolonged); 21% had marked jaundice (bilirubin level > 170 µmol/L [10 mg/dL]); and 18% had abdominal pain. Preexisting underlying disease alone or in combination was present in 17% of patients; this was alcoholism in 24 patients (9%), diabetes in 16 (6%), chronic hepatitis B or C in 10 (4%), and HIV infection in 4 (2%). No association was seen between the presence of underlying disease and the occurrence of a serious complication (P > 0.2).

The clinical complications encountered are shown in Table 2. Thirty-five patients (14%) had 39 serious complications; 20 of these complications were extra-hepatic and 19 were hepatobiliary. Most were present on admission or were evident within 2 days after admission. Twenty-five percent of patients 40 years of age or older and 11% of patients younger than 40 years of age had at least one complication (P = 0.014). Of the 35 patients with serious complications, only 2 (5.7%) would have been advised to receive vaccine under current guidelines (1 had chronic hepatitis C and 1 was an active user of injection drugs) [1].

Ten patients had findings consistent with acalculous cholecystitis, but none required cholecystectomy. Three patients had prolonged cholestasis with jaundice lasting up to 20 weeks. A 3-year-old child presented with hypoglycemia, hyperammonemia, and mental status changes that resolved with supportive treatment.

Ten patients had evidence of hemolysis and 3 patients had acute renal failure; 2 of the 3 required hemodialysis but recovered. Four patients were known to be pregnant. Two of the 4 presented in the first trimester, and the other 2 presented in the third trimester and had preterm labor with premature delivery. Both mothers and babies recovered. Two patients presented with new-onset diabetes with ketoacidosis. A 14-year-old girl developed pericardial and pleural effusions that required pericardiocentesis for cardiac tamponade.

Five patients died (Table 2). Two of the five died of complications of fulminant hepatic failure; one was awaiting liver transplantation at the time of death. This 28-year-old man developed hepatic coma within 72 hours and herniated from refractory increased intracranial pressure. The other patient who died of complications of fulminant hepatic failure was a 23-year-old man who developed the adult respiratory distress syndrome and died of cardiac arrest. Two patients had prolonged illness with antinuclear antibody titers greater than 1:640 and liver biopsy results consistent with autoimmune hepatitis. Neither of these patients was known to have previous liver disease. One improved clinically while receiving prednisone and azathioprine but died of infectious complications. The other died of gastrointestinal hemorrhage, liver failure, and increased intracranial pressure before immunosuppressive therapy was started. One patient with chronic hepatitis C and a history of ethanol abuse died of progressive liver failure. Three of the 53 patients (5.7%) 40 years of age and older and 2 of the 203 patients (1%) younger than 40 years of age died (P = 0.062).

Discussion

Top Methods Results Discussion Author & Article Info References

Hepatitis A is usually self-limited, and full recovery is seen by 3 months in 85% of cases [5]. Previous reports have shown a case-fatality rate of only 0.01% to 0.03% [6] but an increased risk for death with age [7]. Risk for death is also increased in patients with chronic liver diseases [8]. The financial burden of hepatitis A on the U.S. health care system is substantial [9]; the estimated cost of a recent outbreak of 43 cases was $809 706 [10].

The median age of the patients who died in our series was 40 years. Of the five patients who died, only one had evidence of chronic liver disease. Complications were more common in patients 40 years of age and older, and mortality rates also seemed to be higher in this group. However, two patients younger than 30 years of age died.

Some patients with hepatitis A had unexpected clinical manifestations. Extrahepatic complications occurred in 8% of our patients. Acalculous cholecystitis, hemolysis, acute renal failure, and reactive arthritis have been reported [11-13]. Unusual extrahepatic manifestations included new-onset diabetes, preterm labor, and pericardial and pleural effusions.

We identified two cases of autoimmune hepatitis associated with acute hepatitis A, and at least four similar cases have been reported [14]. The hepatitis A virus may serve as a trigger for autoimmune hepatitis in genetically susceptible persons, a phenomenon that has been reported after rubeola and Epstein-Barr virus infection [15, 16].

Between 1983 and 1992, the Viral Hepatitis Surveillance Program did not identify risk factors for HAV transmission in 40% of cases [17, 18]. We were unable to identify risk factors for transmission in 79% of patients; this points out a limitation of retrospective studies. In addition, study patients were hospitalized at 15 institutions and were cared for by multiple physicians, and the admission history was not uniform. For example, patients who had exposure to asymptomatic children with hepatitis A would not have been documented. The studies performed during hospitalization were not standardized, and some complications may have been overlooked, making our estimates imprecise. Biases related to which patients sought medical care and which physicians chose to hospitalize them may limit the generalizability of our findings. We were unable to assess patients who were not hospitalized.

Immunity to hepatitis A approaches 100% in young adults in developing countries. In contrast, immunity in the United States is low: 38% overall, 11% in persons younger than 5 years of age, and 74% in persons older than 50 years of age [19, 20]. A substantial segment of the U.S. adult population is therefore susceptible to HAV infection, setting the stage for large-scale urban outbreaks. This is exemplified by a recent epidemic in Shanghai that affected more than 300 000 persons and caused 45 deaths [6].

Vaccination has been recommended for persons in high-risk groups, including travelers to areas in which hepatitis A is endemic, persons with occupational exposure, injection drug users, patients with chronic liver disease, homosexual men, and persons in communities with cyclic epidemics [1]. Many persons hospitalized for severe hepatitis A are young and do not have these risk factors. Applying current recommendations for vaccination would not have prevented most of these complications and deaths. Without measures to increase immunity to hepatitis A in young adults, future epidemics with severe consequences, such as the one described here, should be expected in the United States.

Dr. Uhl: Division of Gastroenterology, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160-7350.

Drs. Howard and Williams: Department of Medicine, University of Tennessee, Memphis, 955 Court Avenue, Memphis, TN 38163.

Drs. Riely and Waters: Division of Gastroenterology, Department of Medicine, University of Tennessee, Memphis, 951 Court Avenue, Suite 555D, Memphis, TN 38163.