Forgotten but Not Gone: Learning From the Hepatitis A Outbreak and Public Health Response in San Diego

Abstract

Consider the case of a patient during the recent outbreak of hepatitis A virus (HAV) infections in San Diego County. He was a 63-year-old man with a history of hepatitis C virus (HCV) infection and alcoholic liver disease who presented with a left hip fracture suffered during a fall while intoxicated. He subsequently underwent surgical repair of the fracture. Labs drawn on admission showed a total bilirubin of 0.32 mg/dL, aspartate aminotransferase (AST) of 178 U/L, alanine aminotransferase (ALT) of 69 U/L, and alkaline phosphatase of 90 U/L; results were similar to those a month earlier when the patient presented to the emergency department (ED) for an unrelated issue.

His hospital stay was prolonged because he was homeless. One week after surgery, he developed fever and diarrhea and was started on vancomycin and piperacillin/tazobactam for possible hospital-acquired pneumonia, despite the absence of respiratory symptoms and a clear chest x-ray. The patient also lacked signs or symptoms of alcohol withdrawal to explain the fever. The fever persisted for a week. Cultures were negative, but repeat liver function tests (LFTs) showed a total bilirubin of 3.1 mg/dL, AST above 1,000 U/L, ALT of 792 U/L, and alkaline phosphatase of 276 U/L. Since the HAV outbreak was at its peak, house officers ordered a HAV IgM test, which came back positive, confirming a diagnosis of acute HAV infection. The patient was placed into contact isolation.

Further past medical history revealed that the patient had been diagnosed with HCV infection in 2011, with risk factors that included prior cocaine use and a history of blood transfusions. He suffered from severe alcoholism, having 20 to 40 drinks per week for most of his life. He had numerous prior ED visits for falls, fractures, and intoxication. No formal staging of his liver disease had been performed, but he had no known history of cirrhosis or symptoms consistent with decompensated liver disease. He had a 40-year history of tobacco use, 1 pack per day. He had been homeless for 5 years, living primarily in shelters and eating in soup kitchens; however, he had no known recent sick contacts and reported no recent history of eating raw or undercooked foods.

A week after diagnosis of HAV infection, the patient's LFTs trended back toward normal. However, 2 weeks later, daily fevers returned, along with increased abdominal pain, nausea, increases in LFTs to higher levels than previously found, and an increase in international normalized ratio (INR). Additional laboratory work showed that the patient had a high HAV RNA level during this relapsing phase of the illness. HAV IgG remained negative. Of note, the patient had been found to be IgG negative a year earlier, and despite documentation of lack of immunity to HAV, had not received HAV vaccination.

Liver ultrasound showed hepatomegaly with diffuse fatty liver and possible nodularity, trace ascites, normal spleen, normal common bile duct, and normal portal vein. Liver biopsy showed severely active hepatitis, a fibrosis stage of 1 out of 6, portal inflammation primarily with lymphocytes and macrophages, bile duct injury, no steatosis, and spotty necrosis. The patient became increasingly encephalopathic and developed progressive liver failure; he became comatose, was transitioned to comfort care, and died approximately a month and a half after admission to the hospital.

This patient's case highlights 3 important points. First, he died from a completely preventable condition and his story speaks to the need for improved public health measures such as vaccination among our most vulunerable populations. Second, he had a rare complication of acute HAV infection in the form of a relapsing course. And finally, he had many of the risk factors for both acquiring HAV and having a poor outcome from the infection that were consistently observed during the outbreak in San Diego.

Characteristics of HAV Infection

HAV transmission occurs primarily person-to-person by the fecal-oral route. Stool is infectious 2 to 3 weeks before and 1 week after the onset of symptoms. The peak of infectivity precedes the onset of symptoms. Routes of transmission include food (eg, cooking temperature is not high enough, food is prepared by someone who is shedding virus), blood transfusion, illicit drug use, and sex via oral-anal or digital-anal contact. Humans are the primary natural host with genotypes 1, 2, and 3 causing the majority of infections. Although not tested routinely in clinical practice, genotype 1 is the most common in the United States. Genotypes 1b and 3a have been associated with fulminant hepatitis during prior outbreaks in the United States and Korea, respectively.

Figure 1 shows the incidence of HAV infection in the United States reported by the Centers of Disease Control and Prevention (CDC) between 1980 and 2012. The decline starting in 1996 marks the development and implementation of an effective anti-HAV vaccine. Universal vaccination of children began in 2006. According to CDC data, approximate numbers of cases and death rates in recent years have been fairly stable: 2700 and 2.5% in 2011, 3000 and 2.2% in 2012, 3500 and 2.3% in 2013, and 2500 and 3.0% in 2014.

Open in a separate window

Figure 1.

Incidence of hepatitis A virus infections in the United States, 1980–2012. Adapted from Boston University School of Public Health¹

The asymptomatic incubation period of HAV infection is approximately 1 month (range, 15–50 days). Symptoms include fever, fatigue, nausea, anorexia, abdominal pain, jaundice (in up to 70% of cases), vomiting, diarrhea, dark urine, and light-colored stools. Duration of symptoms is generally less than 2 months, with 10% to 15% of cases having a prolonged or relapsing course. Approximately 20% to 40% of patients require hospitalization. Figure 2 shows the clinical, virologic, and serologic events that occur during the course of infection. After the incubation period, viremia occurs followed by shedding of virus in the feces. With the development of symptoms, increases in transaminases are observed. Next, the host immune response begins, with increases in IgM and a subsequent switch to IgG antibody. Anti-HAV IgG antibody is thought to provide lifelong protection against future infection.

Open in a separate window

Figure 2.

Course of clinical, virologic, and serologic events in hepatitis A virus (HAV) infection. ALT indicates alanine aminotransferase. Adapted from Bennett, Dolin, and Blaser.²

Uncommon complications associated with HAV infection include relapsing hepatitis, as was experienced by the patient described, cholestatic hepatitis, autoimmune hepatitis, and acute liver failure. Relapsing hepatitis occurs in approximately 10% of patients (range, 3%–20%) however the underlying etiology and associated risk factors are unclear. The clinical course is that of improvement in symptoms after initial presentation, followed by an increase in LFTs (with or without symptoms). Relapse is typically milder than the initial presentation. HAV RNA levels are markedly elevated, similar to levels during initial infection. Relapse occurs 1 to 12 weeks after the initial presentation and can persist for 3 weeks to 12 months, with no well-defined sequelae.

Cholestatic hepatitis occurs in approximately 5% of patients and is marked by prolonged (>3 months) elevation of bilirubin and alkaline phosphatase levels. The clinical course is marked by prolonged jaundice, pruritus, fever, weight loss, and diarrhea. It typically resolves spontaneously and is treated with supportive care. Cholestyramine can be given for pruritus.

Acute liver failure is a rare but devastating complication characterized by elevated LFTs (especially bilirubin), co-agulopathy, and hepatic encephalopathy in a patient with previously stable liver function. It is estimated that less than 0.1% of HAV cases result in acute liver failure. Risk factors include older age (>50 years), coinfection with hepatitis B virus (HBV) or HCV, and underlying liver disease.

Treatment and Prevention

Treatment for HAV infection is supportive care, including hydration, nutrition, and avoidance of hepatotoxins. The vast majority of affected individuals completely recover by 6 months following initial infection, usually much sooner.

Vaccination is a highly effective mode of prevention, with 2 inactivated whole-virus vaccines given in 2 doses 6 months apart. A combined HAV and HBV vaccine is also available. Approximately 95% of adults exhibit seroconversion to IgG antibody after the first dose of the 2-dose vaccine, with approximately 100% seroconverting after the second dose. Seroconversion rates are lower in patients with underlying liver disease and in those with advanced immunosuppression. For example, among people with HIV infection, seroconversion rates are reported to range from 50% to 90%. Risk factors associated with lack of seroconversion include low CD4+ cell count, high HIV RNA level, HCV coinfection, and tobacco use. For immunocompetent hosts, modeling studies indicate that greater than 90% of persons maintain protective antibodies 40 years after vaccination. Durability of immunity in immunocompromised host thought to be lower but is less well characterized.

The current indications for vaccination include all children at 1 year of age, travel to endemic countries, chronic liver disease (eg, HBV or HCV infection, cirrhosis), clotting factor disorders, men who have sex with men (MSM), illicit substance users, and persons working with non-human primates. Other groups, often considered for vaccination indications, but in whom vaccination currently is not required outside of the setting of outbreaks, include those at higher risk of acquiring and transmitting HAV during outbreaks— eg, healthcare workers, food service workers, and people who work with children, such as in daycare settings. Pre-vaccination serologic testing is not recommended in persons with an indication for vaccination. Post-vaccination testing for response is not indicated in immunocompetent hosts; its role in imunocompromised persons is unclear.

Postexposure prophylaxis, consisting of the HAV vaccine series and immune globulin in most cases, should be given to non-immune individuals as soon as possible and within 2 weeks after exposure to a confirmed case of HAV infection. The recommended dose of immune globulin is now 5 times higher than that previously recommended, because more recent preparations were found to have lower concentrations of HAV antibodies, likely reflecting the lower prevalence of HAV in plasma donors in recent years.

San Diego Outbreak

Figure 3A shows the number of cases of HAV infection in San Diego between 1994 and 2018, highlighting the 2017 outbreak; Figure 3B shows the onset of outbreak cases by week. There were 590 confirmed outbreak cases between November 22, 2016, and June 21, 2018, all involving HAV genotype lb. Among the cases, 402 (68%) were boys or men, with 14 being MSM. Among all patients, 405 (69%) were hospitalized and 20 (3.4%) died. This mortality rate was substantially higher than in prior years as outlined above. The median age of the cases was 43 years (most age 40 to 60 years; range, 5–87 years). Risk factors for infection included homelessness and illicit drug use in 194 cases (34%), homelessness alone in 91 (15%), and illicit drug use alone in 77 (13%), with 167 (28%) being associated with neither of these risk factors. The risk factors in 56 (10%) were unknown. Geographically, the outbreak was centered in areas with high rates of homelessness. Approximately 17% of patients had HCV coinfection and approximately 5% had HBV coinfection. In prior years, most cases of HAV infection in San Diego were associated with international travel.

Open in a separate window

Figure 3.

A: Hepatitis A virus infection cases in San Diego, California, 1994 to 2018. B: Number of cases per week in 2017 outbreak. Courtesy of the County of San Diego, Health and Human Services Agency, Public Health Services, Epidemiology & Immunization Services.

The outbreak spread beyond San Diego, with cases genetically and epidemiologically linked to the San Diego strain including 76 in Santa Cruz, 12 in Los Angeles, 12 in Monterey, and 17 in other areas in California. Linked cases were also identified in Arizona, Colorado, Kentucky, Indiana, Rhode Island, West Virginia, and Utah.

The outbreak in San Diego and an ongoing outbreak in Michigan underscore a change in the nature of HAV outbreaks from mostly food-borne to centered on homelessness, sanitation, and illicit drug use. For example, one of the largest post-vaccine era outbreaks occurred in 2003 in association with green onions at a restaurant chain in Pennsylvania and Ohio; a total of 660 cases were identified, with 3 deaths. A more-recent food-borne outbreak in 2016 in Hawaii, associated with raw scallops from a sushi restaurant, involved 292 cases, 74 hospitalizations, and no deaths.

In contrast, the southeast Michigan outbreak, ongoing since 2016, shares features with the San Diego outbreak, including male predominance, high morbidity and mortality, association with illicit substance use and homelessness, and cases with HCV coinfection. Thus far, there have been 850 cases, with a median age of 40 years and 65% being male individuals. Hospitalization has been required in 80% of cases and 3.2% have died. More than 50% of cases were associated with illicit substance use and 13% with homelessness; 27% of cases had coinfection with HCV. Understanding the changing epimediology of hepatitis A transmission and outbreaks will be important with regard to prevention efforts going forward.

San Diego Public Health Response

In response to identification of the HAV outbreak, a local health emergency was declared on September 1, 2017. This was followed by a declaration of a state of emergency by the governor of California on October 13, 2017. A public health strategy of vaccination, sanitation, and education was instituted by the local health department.

The vaccination effort included administration of more than 160,000 vaccinations, with 85% administered to high-risk individuals. More than 30,000 vaccinations were given in the field through mass vaccination events, mobile foot teams, and mobile vans. The public health department partnered with local health systems (eg, University of California San Diego, Scripps, Sharp, Kaiser, Family Health Centers), implemented outreach to those at highest risk (eg, homeless persons and others who lacked health insurance or regular contact with the health system), and implemented a program for vaccination in EDs. The routine indications for vaccination were expanded to include: all food service workers; healthcare workers, sanitation workers, and public safety workers who work with at-risk populations; homeless service providers and volunteers; and any person who desired immunity.

The public health strategy was devoted to the effort of going to the streets to meet vulnerable persons and individuals at higher risk to provide education about the outbreak and to offer and provide vaccination. As noted, the initiatives included holding staffed vaccination events at homeless shelters. Mobile vans traveled from place to place, focusing on downtown San Diego and El Cajon, where most of the cases were occurring, with personnel providing education and vaccinations to people who were willing to receive them. Mobile foot teams would go from tent to tent and person to person on the street in areas with high concentrations of homeless individuals, again providing education about the outbreak and offering and providing vaccination. Figure 4 shows the outcome of these efforts, with the number of new infections dropping once the vaccination effort reached its peak.

Open in a separate window

Figure 4.

Effect of vaccination campaign in controlling San Diego hepatitis A virus outbreak. Courtesy of the County of San Diego, Health and Human Services Agency, Public Health Services, Epidemiology & Immunization Services.

The sanitation campaign included the establishment of handwashing stations (not employing alcohol-based sanitizers, which are not effective), portable toilets and increased access to public restrooms, and hygiene kits, particularly in the downtown areas. Food inspections were also conducted, particularly during the early part of the outbreak when food-borne transmission was more highly suspected. As part of the sanitation campaign, the city employed power spraying of sidewalks to remove fecal matter that accumulated in the setting of homelessness and inadequate access to bathroom facilities. The city also built 3 large industrial tents that could house 700 people as a temporary housing option for people living on the street, an initiative that cost $6.5 million. The tents were staffed with 24-hour security, contained bathrooms and showers, and were staffed by personnel to provide services during the day that included mental health and substance abuse services, job training and case management, and bridging to permanent housing.

As part of the education campaign, the department of public health provided community presentations and community outreach and staffed a 24-hour hotline to dispense information about vaccination events. Broad media coverage of the outbreak was provided to augment public awareness. Local and state political will to battle the outbreak was mobilized and played a large part in securing the resources needed for the public health efforts to succeed.

Summary

The vaccination, education, and sanitation campaigns in the San Diego outbreak were very robust. The strategies employed were innovative and highly individualized in an attempt to ensure vaccination of those at highest risk. The question remains, however: is this enough?

Homelessness and its association with poor sanitation constituted a root cause of the outbreak. Homelessness was also associated with many challenges in trying to stop the outbreak that differ markedly from those encountered in food-borne outbreaks. Although measures to address homelessness were employed in combating the outbreak (eg, the temporary housing with industrial tents), it is clear that much more needs to be done in addressing the homelessness crisis in San Diego and in many areas throughout the United States. Homelessness and associated poor sanitation threaten to make outbreaks of HAV and other infectious illnesses more common in the future. Attention to this threat and the political will to address the crisis need to be heightened if we are to make any substantial progress in preventing such outbreaks.

In summary, the San Diego HAV outbreak is one of the largest outbreaks reported in the United States to date. It was associated with a high degree of morbidity and mortality. Homelessness and substance use were important risk factors identified early on and guided the public health response. Vaccination was the most important strategy to stop the outbreak. There is a continuing need to address the problem of homelessness and other social determinants of health.

Footnotes

References

Additional Suggested Reading