During 2003–2012, an estimated 23,266 hospitalizations (95% CI 21,741–24,792) in the United States were assigned an ICD-9-CM code of 123.1 in any of the first 15 diagnostic fields. Of these hospitalizations, 18,584 (95% CI 17,322–19,846), approximately 80% of the total, met our case definition of hospitalization due to neurocysticercosis. The number of hospitalizations due to neurocysticercosis per year ranged from a high of 2,247 in 2006 to a low of 1,495 in 2012. The largest proportion of hospitalizations due to neurocysticercosis occurred in the western region (n = 8,026, 42.9% [95% CI 39.2%–46.7%)], followed by the southern region (n = 5,860, 31.8% [95% CI 28.6%–35.1%]), the northeastern region (n = 2,902, 15.5% 95% CI [13.5%–17.6%]) and the midwestern region (n = 1,796, 9.8% [95% CI 8.2%–11.7%]).

We found distinct differences in the mean annual incidence rates of hospitalization stratified by age, sex, and race (Table 2). The mean annual incidence of hospitalization was highest in 20- to 44-year-old age group (1.04 hospitalizations/100,000 population). Hospitalization rates were 33% higher among male patients than female patients. The age- and sex-adjusted mean annual incidence of hospitalizations was highest among Hispanics (2.50 hospitalizations/100,000 population); the rate was 35 times higher than that for non-Hispanic whites, 10 times higher than that for blacks, and 8 times higher than that for Asian/Pacific Islanders. Unadjusted rates by race were similar: Hispanic, 2.57/100,000; white, 0.06/100,000; black, 0.23/100,000; and Asian/Pacific Islander, 0.26/ 100,000.

The study findings demonstrate that neurocysticercosis poses considerable health and economic problems in the United States, especially among the Hispanic population. Over the 10-year study period, >18,500 hospitalizations for neurocysticercosis occurred, totaling hospital charges of >US $908 million, of which 40% was billed to publicly funded insurance programs. Hospitalization stays were prolonged and expensive, reflecting the complicated nature of acute disease management. Hospitalizations and associated charges for cysticercosis exceeded the totals for malaria and for all of the other NTDs combined.

The hospitalization rates we report in this nationwide study are comparable to those reported in previous state- or county-level studies, providing support for the case definition we used (12,15–20). Because there is no ICD-9-CM diagnostic code specific for neurocysticercosis, the case definitions varied slightly among these studies. The estimated overall hospitalization rate of 0.65/100,000 population that we report falls between the rates previously observed in California (0.8–1.1 hospitalization/100,000 population) and Oregon (0.2–0.5 hospitalizations/100,000 population) (12,15–18). Risk for hospitalization was highest among Hispanic, male, and young to middle-aged adult patients in all studies.

Nearly three quarters of all patients hospitalized for neurocysticercosis in the United States were Hispanic. The Hispanic population is the largest minority group in the United States and among the fastest growing US population groups. Because the hospitalization rate for the Hispanic population is 36 times greater than that of the non-Hispanic white population, the effect of neurocysticercosis on the US economy is likely to increase substantially in the coming years. The US Census Bureau projects that the Hispanic population will grow from 53 million in 2012 to >78 million by 2030 (21). Without changes in the rate of hospitalization or the increase in mean hospitalization charges, there could be >1,900 hospitalizations and US $250 million total charges related to neurocysticercosis among Hispanics alone in the year 2030. Changing immigration patterns may also bring an influx of cases in persons from other regions of the world where neurocysticercosis is endemic, particularly Asia and sub-Saharan Africa.

Several hospital-based studies have shown that seizures are the most frequent reason for hospitalization for neurocysticercosis (3,4,22). In this study, epilepsy was the most frequent diagnosis associated with hospitalization for neurocysticercosis; it was coded in more than half of all hospitalizations for the disease. Seizures in neurocysticercosis are typically amenable to therapy with antiepileptic and anti-inflammatory drugs, resulting in relatively uncomplicated and short hospital stays. In contrast, more severe disease may require intensive interventions and longer hospitalizations, resulting in higher charges (23–25). While diagnoses of obstructive hydrocephalus or encephalitis/meningitis occurred in ≈20% of persons hospitalized for neurocysticercosis, these more severe presentations accounted for 40% of the total charges incurred.

We report hospitalization diagnostic codes that may not represent the distribution of disease manifestations experienced by individual patients. For example, although a diagnostic code for headache was listed for 11% of hospitalized patients, only patients with headaches associated with underlying pathology requiring acute intervention, such as obstructive hydrocephalus, are likely to be admitted and therefore represented in this study. Even then, the diagnosis of headache may be underrepresented. There were twice as many hospitalizations with diagnostic codes for hydrocephalus and encephalitis than for headache, although both of these manifestations would be expected to be associated with headache (22). Similar caution is suggested in interpreting the frequency of other diagnoses presented here. It may seem contradictory that only 17% of hospitalizations had a procedural code for neuroimaging. However, because most imaging for neurocysticercosis would be expected to occur in the emergency department before admission, the infrequent coding for neuroimaging may reflect exclusion of these procedural codes from the hospital discharge summary.

This study documents the substantial costs of hospitalizations due to neurocysticercosis in the United States, but the true effect of neurocysticercosis on the US health care system is likely much greater. Only those emergency department visits that result directly in inpatient admission are captured in the hospital discharge databases in the NIS. In Oregon (15), over 40% (31/72) of all patients with neurocysticercosis were seen only in the emergency department and were not admitted to the hospital. While nonadmissions likely represent cases of less clinical severity, substantial charges are still incurred in the emergency department and in outpatient follow-up. Neurocysticercosis is also likely to be substantially underdiagnosed and misdiagnosed because of the lack of a definitive diagnostic test and limited provider awareness of the disease.

Neurocysticercosis also often results in chronic disease that requires outpatient follow-up with infectious disease or neurology specialists, none of which is captured in this study. Management of neurocysticercosis may involve long-term antiepileptic therapy, prolonged regimens of antiparasitic drugs and high-dose corticosteroids, monitoring and repair of ventriculoperitoneal shunts, and treatment of frequent complications resulting from these interventions (26,27). A chart review at the outpatient neurology clinic in a Houston hospital showed that 2% of all patients were seen for management of neurocysticercosis (28). A few states are now collecting comprehensive claims data covering health care provided in inpatient, outpatient, and long-term care settings. Data from these programs could provide more complete information about health care and associated costs related to management of neurocysticercosis in all settings. The high neurocysticercosis hospitalization rate we noted in young adults and men suggests substantial indirect costs to the US domestic workforce. Loss of worker productivity should also be considered in the overall costs of neurocysticercosis.

The use of administrative databases, such as the NIS, to obtain data for this study does have drawbacks, including several limitations we already described. An additional drawback to using the NIS was the inability to identify multiple hospitalizations for a single person, which precludes the ability to estimate the prevalence or incidence of disease. Although the number of states participating in NIS has grown over the years, several states still do not participate in reporting hospital discharge data. In addition, the regional sampling structure of the NIS does not allow for accurate state-level estimates, limiting the ability to identify specific states whose populations are at increased risk for neurocysticercosis. Furthermore, the lack of in-depth demographic and clinical information in the NIS limits the type of questions that can be addressed. For example, knowing the country of birth or travel history of patients with neurocysticercosis could help understand their source of exposure.

Although the primary purpose of this study was to evaluate hospitalizations for neurocysticercosis, we also compared hospitalizations for cysticercosis with those for other NTDs and malaria. Our findings showed that the number of hospitalizations for cysticercosis was nearly 2 times the number for malaria, and the associated hospital charges were nearly 3 times higher. In addition, hospitalizations and charges for cysticercosis were higher than those for all other NTDs we evaluated combined. This comparative analysis was not meant to be exhaustive; we recognize that many factors other than hospitalization contribute to the public health effect of any particular disease. However, the markedly higher number of hospitalizations and associated charges related to cysticercosis, compared with those for other NTDs and malaria in the United States, merits attention and further exploration.