Cryptosporidium Testing Practices Among Clinical Laboratories in the United States

Cryptosporidium is increasingly recognized as an important diarrheal pathogen, especially in persons with compromised immune systems. In immunocompetent persons illness is usually self-limited. Symptoms include watery diarrhea and abdominal cramps; nausea, vomiting and low grade fever may also occur. Patients with immunodeficiencies may develop voluminous diarrhea, resulting in dehydration, electrolyte disturbances and death.1 Because Cryptosporidium is usually detected by microscopic examination, clinicians might assume that stools submitted for ova and parasite testing are examined for the presence of Cryptosporidium oocysts. However, identification of Cryptosporidium oocysts requires special staining or other techniques that many laboratories may not perform routinely.2 We conducted a nationwide survey of clinical laboratories to determine the percentage that test for Cryptosporidium and the procedures they use for testing. Methods. In December, 1994, and January, 1995, we surveyed a sample of US laboratories by randomly selecting 230 (2%) of the 11 548 laboratories in the US that have applied for certification to perform high or moderately complex tests as defined by the Clinical Laboratory Improvement Amendments of 1988. We conducted telephone interviews with the microbiology supervisor of each laboratory. For analysis states were divided into 4 standard regions: Northeast, Midwest, South and West. We used Fisher's exact test to compare differences among groups. For continuous variables, analysis of variance was used to compare the means. All P values are two tailed. Results. We were able to contact 229 of the 230 selected laboratories. Of these 129 (56%) tested stools for microbiologic organisms and thus were eligible to participate; 117 (91%) were hospital-based and 12 (9%) were independent laboratories. The median annual volume of all tests performed in these laboratories was 187 765 (range, 700 to 5 204 646). Thirty-six states were represented. Of the 129 laboratories that performed microbiologic testing of stool specimens, 94 (73%) tested stools for the presence of ova and parasites. Five (5%) of these laboratories reported testing all stools submitted for ova and parasite examination for Cryptosporidium, and an additional 74 (79%) reported testing at least some stools for Cryptosporidium. Of these laboratories the most common indication for testing was physician request; all 74 laboratories reported testing when requested by the physician; 35 laboratories tested only on physician request (Table 1). Indications for screening did not vary significantly by region of the country. Hospital-based laboratories (87%) were significantly more likely than independent laboratories (50%) to provide testing for Cryptosporidium (P = 0.02). Laboratories with greater annual test volume were also more likely to test for Cryptosporidium (P = 0.07). We asked the microbiology supervisors of the 79 laboratories that reported screening for Cryptosporidium what method or methods they routinely used to test stool specimens for the organism. Sixty-one (77%) reported using a modified acid-fast stain, 19 (24%) used a monoclonal fluorescent antibody test and 9 (11%) used auramine-rhodamine stain; none reported using an enzyme-linked immunosorbent assay. Discussion. Since its initial description as a human pathogen in 1976,3, 4Cryptosporidium has become an increasingly recognized cause of diarrheal illness, especially in immunocompromised persons. Recently large waterborne and foodborne outbreaks of cryptosporidiosis have brought greater attention to the organism as a cause of gastroenteritis in immunocompetent hosts.5-7 Because the infectious dose is as small as 30 oocysts,8 person-to-person transmission is also common, especially in day-care centers.9 The incidence of sporadic infections in the United States is unknown, and lack of testing and reporting contributes to its underrecognition as a cause of diarrhea. A study in Canada of all stools submitted for microbiologic examination found that Cryptosporidium oocysts were detected in fecal samples of 1.2% of patients; among parasites only Giardia was detected more frequently.10 In England a large, multicenter study of patients with diarrhea found that infection with Cryptosporidium was nearly three times more common than with Shigella.11 Clinicians should be aware that even among laboratories that perform ova and parasite testing, most laboratories do not routinely screen stool specimens for Cryptosporidium. When they do test for Cryptosporidium three-fourths of laboratories use modified acid-fast staining. The sensitivity of this stain for Cryptosporidium is relatively low, 41 to 76%, particularly in cases in which the diarrheal illness is mild.2, 12 Auramine-rhodamine staining may be more sensitive but lacks specificity.12 Enzyme-linked immunosorbent and monoclonal antibody techniques appear to be have the best sensitivity and specificity but are more costly.2, 12, 13 Reasons for laboratories not routinely testing for Cryptosporidium were not assessed by our survey. Possible reasons include the perception that the local incidence is low, concerns about the cost of testing and the fact that there is currently no proven therapy. A study in Canada in 1988 found that depending on the method used, the cost of testing ranged from $4.45 to $7.96 per specimen.2 A 1994 survey of laboratories licensed to test for fecal parasites in Connecticut found that the median additional charge for adding a Cryptosporidium test to an ova and parasite examination was $29 (range, $0 to $116).14 Although there is still no proven antimicrobial therapy for cryptosporidiosis, making the diagnosis can save costs by obviating the need for further diagnostic tests. In addition the public health benefits of screening for Cryptosporidium are great. Without screening, small outbreaks are likely to go undetected7 and large outbreaks may be detected less quickly,5 resulting in many additional cases. Early detection of outbreaks is especially important in day-care centers, where rapid intervention can prevent secondary spread.9 Making the diagnosis of cryptosporidiosis is difficult and requires a high index of suspicion. Although no symptoms are pathognomonic laboratory-confirmed illness is most frequently characterized by watery diarrhea and abdominal cramps.5 Until laboratory screening for Cryptosporidium becomes routine, clinicians should specifically request testing for this pathogen when the diagnosis is suspected. Acknowledgments. We thank Health Care Direct, Inc., Flemington, NJ, for supplying the list of laboratories; Christine Roberts, M.B.B.S., M.P.H., and the Connecticut Department of Public Health and Addiction Services for assistance with developing the questionnaire; Dennis Juranek, D.V.M., for reviewing the manuscript; and Lynne McIntyre, for editorial assistance. Thomas G. Boyce, M.D.; Anne G. Pemberton, M.D.; David G. Addiss, M.D., M.P.H.