Parasitology in 2018: Should we still be ordering O&P times three?

A terrific review article (S Mohapatra et al. Am J Gastroenterol 2018; 113: 805-18) provides a great deal of information about gastrointestinal parasites. Thanks to Ben Gold for this reference (& don’t forget to vote for NASPGHAN president).

Generally, the authors dispute the usefulness of testing for ova and parasites (O&P) with three separate specimens.  While classic training has noted the intermittent shedding of parasites and the suboptimal sensitivity of O&P, the authors note that a recent study showed a detection of 91% of parasites in the first stool sample.  In addition, newer PCR based assays are more appropriate in many clinical situations due to their improved sensitivity.

The authors first review the protozoa, which are single-celled, motile, free-living organisms, in depth & summarized in Table 1; these include the following:

  • Amoeba: Entamoeba histolytica (E histolytica),
  • Dientamoeba fragilis
  • Blastocystis hominis
  • Coccidia: Cryptosporidium, Cystiospora, Cyclospora
  • Ciliates: Balantidium coli
  • Flagellates: Giardia lamblia
  • Microsporidiosis
  • Trypanosoma cruzi

Next, they review the helminths in depth and in Table 2, which are large, multicellular organisms that can be seen with the naked eye and include the following:

  • Ascariasis: A lumbridcoides
  • Capillariasis
  • Diphyllobothriasis
  • Enterobiasis: E vermicularis
  • Hookworm disease: A dudenale, N amercanus
  • Hymenolepiasis
  • Strongyloides: S stercoralis
  • Schistosomiasis
  • Taeniasis
  • Trichinellosis
  • Trichuriasis
  • Groups of helminths: trematodes (eg. Schistosomes), cestodes (tapeworms eg. Taenia), and nematodes (roundworms eg. Ascariasis, hookworm, pinworms, and whipworms).

Key points:

  • For E histolytica, ELISA fecal antigen test is superior to O&P as is the PCR assay.  If the diagnosis of E histolytica is being considered in the setting of ulcerative colitis, the authors note that this infection must be excluded before the initiation of corticosteroid therapy since steroids can lead to hyperinfection and could be fatal.  Also, the so-called “flask shaped” ulcers seen with this infection refers to the microscopic appearance of the ulcer into the submucosa. Most infections (>90%) remain asymptomatic.
  • Blastocystis “is the most common parasite identified in stool samples in the US” though the pathogenicity remains controversial and is often self-limited.
  • D fragilis “as a pathogen is controversial…[but] recent studies on patients infected only with D fragilis have found an association with diarrhea, abdominal pain, nausea, weight loss, anorexia, and flatus which resolve after eradication.”
  • Giardiasis is “the most common intestinal parasitic disease affecting humans in the US.” PCR/molecular methods are highly sensitive (>90%) and specific (nearly 100%)
  • Enterobius vermicularis (pinworms). The “CDC does not recommend stool examination for O&P since the yield is low.” The diagnostic test is the “Scotch test” in which tape is left overnight in the perianal region and then examined for captured eggs.

Author Recommendations:

  • “Restrict stool examination [for parasites] to patients with persistent diarrheal illness with a duration greater than 7 days.”  Do not check O&P in hospitalized patients more than 3 days into their hospitalization.
  • The most  common parasitic infections, Giardia and Cryptosporidium, are best diagnosed with a stool immunoassay (EIA) rather than O&P.  For E histolytica EIA is recommended over O&P.
  • In those who are persistently symptomatic and with travel history with likely parasite exposure, stool O&P with wet mount/AFB stain/special stains for detection of rare parasites still is worthwhile.  In those without exposure history and with persistent diarrhea (after exclusion of Giardia and Cryptosporidium), consider non-infectious causes of diarrhea.
  • We discourage repeating the O&P due to the “very low incremental yield of second and third samples”

My take: This article makes a strong argument that “O&P times three” represents an outdated approach in the diagnosis of parasitic diseases in the US.

Related blog posts:

Near the top of Old Rag Mountain, Shenandoah Natl Park




Does it make sense to look for parasites in RAP?

Probably not, in the absence of other symptoms.  A small study from Switzerland shows that treatment of Blastocystis hominis does not help recurrent abdominal pain (RAP) more than placebo (JPGN 2012; 54: 677-79).

This study result is not surprising.  A previous study of 157 RAP pediatric patients (JPGN 2007; 44: 524-26) showed that testing for ova and parasites was unnecessary as the incidence of parasitic infections in this population was not increased in RAP patients compared with controls. In this 2007 study, stool samples were positive for parasitic infection in 15 children, with no difference in prevalence between children with chronic abdominal pain (6/87; 7%) and healthy control children (9/70; 13%).

The current study had 37 patients complete the study.  The pain index (PI) for all patients improved in the study; however, there was no difference between patients who received antibiotics (trimethoprim-sulfa was used as first-line agent) compared to those who received placebo.

Regarding the methodology, the PI was either a visual analog scale with numbers from 0 to 10 or a standardized face scale; the specific scale was determined by the patient age.  Also, among the 20 patients who received antibiotics, 13 continued with B hominis in their stool tests after an initial round of therapy.  Yet the change in pain was virtually identical.  In antibiotic-treated patients, initial PI was 7.1.  After treatment, those with persistent B hominis detected in the stool had a PI of 4.2 whereas those with negative stool specimens had a PI of 4.1.  The placebo patients started with a PI of 7.4; after placebo treatment, a similar proportion cleared the B hominis from their stool –in this group, at the conclusion of the placebo treatment the PI was 2.4; the placebo group with persistent detection of B hominis had a PI of 3.2 at the conclusion of the study.

Although the potential pathogenicity of B hominis in humans remains unclear, it is not unusual in clinical practice for patients with this finding to receive treatment.   Besides trimethoprim-sulfa, metronidazole is frequently administered. Based on this study, placebo would be as helpful as an antibiotic in improving clinical symptoms.

Related blog entries:

What to make of FODMAPs

Pain changes brain