A Guide to Utilization of the Microbiology Laboratory for Diagnosis of Infectious Diseases: 2018 Update by the Infectious Disease Society of America and the American Society of Microbiology1 includes guidance for infections of the gastrointestinal tract. The guide states for many gastrointestinal infections, particularly noninflammatory diarrhea and acute gastroenteritis of short duration, no laboratory testing is recommended. The specimen of choice to diagnose diarrheal illness is diarrheal stool, not a formed stool or a swab, except in pediatrics where a swab is acceptable when feces are noted on the swab. Multiple stool specimens are rarely indicated for the detection of stool pathogens. Toxin or nucleic acid amplification testing for C. difficile should only be performed on diarrheal stool unless the patient has a confirmed diagnosis of ileus.1
The guide states the appropriate approach to the diagnosis of diarrheal illness is determined by the patient’s age and status, severity of the disease, duration and type of illness, time of year, and geographic location. For severe, bloody, febrile, dysenteric, nosocomial, or persistent diarrhea, fecal testing using culture or culture-independent methods is indicated. To determine what organisms, methods, and screening parameters are included as part of the routine culture or culture-independent method, communication with the laboratory is required. Stool cultures often fail to detect the causative agent; therefore when necessary, culture-independent methods are recommended as adjunct methods.1
Culture-independent methods can detect pathogens in one to five hours compared to 24 to 96 hours required for cultures. The assays are reported to be more sensitive than culture and have much higher detection rates. Highly multiplexed assays allow for the detection of mixed infections. The importance of detection of each pathogen is unclear. The clinical significance of a greater number of pathogens detected by the more sensitive assay is uncertain, and could confound treatment. Culture-independent methods should not be used as a test of cure, because they will detect both viable and nonviable organisms.1,11
C. difficile toxin detection by either enzyme immunoassay (EIA) or immunochromatographic methods are widely used in clinical practice. These tests have a reported sensitivity of 70%-85% with faster result time than the toxigenic culture and the cytotoxin assay. Glutamine dehydrogenase antigen assays are sensitive but not specific. NAATs for the detection of C. difficile have a reported sensitivity of 93%-100%.1 NAATs detect viable and nonviable organisms; therefore, to reduce the identification of colonized patients, some laboratories are performing both NAATs and toxin production tests. When testing is limited to patients with unexplained and new-onset diarrhea who are not receiving laxatives, NAAT alone or toxin EIA as part of a multistep algorithm are recommended.
Historically, it has been recommended that three specimens collected over a seven-to-ten-day period be submitted for ova and parasite (O&P) examination. Options today include O&P examination of a second or third specimen if the previous was negative and the patient remains symptomatic. Targeted use of immunoassay testing or NAAT for the most common parasites based on geographic location, patient demographics, and physician request can be used as a screen, only for negative patients with continued symptoms or for patients with specific risk factors requiring full O&P examination. Immunoassays for Giardia are sensitive enough that only one specimen may be needed. No data is currently available on the number of specimens required to rule out infection via NAAT. Pathogenic Entamoeba histolytica can only be differentiated from nonpathogenic Entamoeba dispar using an immunoassay or NAAT.1
Viral gastroenteritis is often of short duration and self-limited, and viral shedding may persist after resolution of symptoms. Testing via multiplex NAATs is not routinely performed except in immunocompromised patients, for infection control purposes, or for outbreak investigations. Testing for cytomegalovirus in immunocompromised patients should be performed using a quantitative NAAT performed on plasma.1
The Clinical Practice Guidelines for Clostridium difficile Infection in Adults and Children: 2017 Update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA)11 states their recommendations are intended to optimize patient care based on a systematic review of the evidence and an assessment of the benefits and harms of alternative care options. Per the guideline, repeat testing (within seven days) during the same episode of diarrhea is not recommended. Also, stool testing from asymptomatic patients is not recommended, except for the purpose of epidemiological studies (strong recommendation, moderate quality of evidence).11
The American College of Gastroenterology (ACG) Clinical Guideline: Diagnosis, Treatment, and Prevention of Acute Diarrheal Infections in Adults5 focuses primarily on immune-competent adult individuals and does not consider C. difficile associated infections. For epidemiology and public health considerations, diagnostic evaluation using stool culture and culture-independent methods, if available, should be used when the individual patient is at high risk for spreading the disease and during known or suspected outbreaks. The recommendations for diagnosis include use of stool diagnostic studies, if available, in cases of dysentery, moderate-to-severe disease, and symptoms lasting greater than seven days in order to clarify the etiology of the illness and enable specific directed therapy. For the majority of cases of acute diarrheal illness, traditional methods of diagnosis fail to identify the etiology. If available, the use of Food and Drug Administration approved culture-independent methods of diagnosis can be recommended at least as an adjunct to traditional methods. Antibiotic sensitivity testing for management of acute diarrheal infections is not recommended.
While not considered in the ACG guidelines5, the work-up for chronic diarrhea is briefly discussed and recommended to include the differential diagnoses such as celiac disease, Crohn’s disease, eosinophilic gastroenteritis, and Whipple’s disease. In the situation of chronic diarrhea occurring after infectious diarrhea, a diagnosis of postinfectious IBS must be considered.10 The likely pathogens in cases of infectious chronic diarrhea are the parasites Cryptosporidium spp, Giardia lamblia, Cyclospora cayetanensis, Cystoisospora belli, and Entamoeba histolytica.10
The 2017 Infectious Diseases Society of America Clinical Practice Guidelines for the Diagnosis and Management of Infectious Diarrhea10 recommend a detailed clinical and exposure history be obtained from patients with diarrhea under any circumstance with adherence to jurisdictional recommendations for outbreak reporting and infection control. Patients with fever or bloody diarrhea should be evaluated for enteropathogens for which antimicrobial agents may provide clinical benefit. Enteric fever should be considered in a febrile patient with exposure to Salmonella enterica subspecies Typhi or Paratyphi. When a possible Shiga toxin-producing organism is suggested by clinical or epidemic history, diagnostic approaches should be applied that detect Shiga toxin and distinguish Escherichia coli (E. coli) O157:H7 from other Shiga toxin-producing E. coli (STEC) in stool. Diagnostic tests that can distinguish between the more potent Shiga toxin 2 and less potent Shiga toxin 1 could be used. Shigella dysenteriae type 1, and rarely other pathogens, may produce Shiga toxin and should be considered as a cause of hemolytic-uremic syndrome (HUS), especially in patients with a recent history of international travel or personal contact with a traveler. Clinicians should evaluate for postinfectious and extraintestinal manifestations of enteric infections.
Stool testing should be performed to detect Salmonella, Shigella, Campylobacter, Yersinia, C. difficile, and STEC in patients with diarrhea who also have fever, bloody or mucoid stools, severe abdominal cramping or tenderness, or signs of sepsis. Bloody stools are not expected with C. difficile infections. STEC O157 should be assessed by culture and non-O157 STEC should be detected by Shiga toxin or genomic assays. Specifically, test for Yersinia enterocolitica in patients with persistent abdominal pain and fever at epidemiologic risk for yersiniosis. Stool specimens should be tested for Vibrio species in patients with large volume rice water stools or exposure to either salty or brackish waters, who have consumed raw or undercooked shellfish, or who have traveled to cholera-endemic regions within three days prior to onset of diarrhea. A broader set of bacterial, viral, and parasitic agents should be considered in the context of a possible outbreak of diarrheal illness. Selection of agents for testing should be based on a combination of host and epidemiologic risk factors and ideally in coordination with public health authorities.10
For immunocompromised patients with diarrhea, a broad differential diagnosis is recommended for evaluation of stool specimens by culture, viral studies and examination for parasites. Patients with acquired immune deficiency syndrome and persistent diarrhea should undergo testing for additional organisms including, but not limited to, Cryptosporidium, Cyclospora, Cystoisospora, microsporidia, Mycobacterium avium complex, and CMV.10
Clinical consideration should be included in the interpretation of multiplex NAAT because these assays detect deoxyribonucleic acid (DNA) and not necessarily viable organisms. All specimens that test positive for bacterial pathogens by culture-independent diagnostic testing such as antigen-based molecular assays (gastrointestinal tract panels), and where isolate submission is requested or required under public health reporting rules, should also be cultured in the clinical laboratory or at a public health laboratory to ensure that outbreaks of similar organisms are detected and investigated. A culture may also be required when antimicrobial susceptibility testing results would affect care or public health responses. Specimens from patients involved in an outbreak should be tested per public health department guidance.10
In cases of outbreak when there is clinical suspicion of enteric fever or diarrhea with bacteremia, it is recommended that culture independent testing, including molecular diagnostics, should be performed on stool and blood specimens.10
Testing may be considered for C. difficile in patients older than two years of age, with a history of diarrhea following antimicrobial use, or in patients with healthcare associated diarrhea. A single diarrheal stool specimen is recommended for detection of toxin or a toxigenic C. difficile strain.10
A diarrheal stool specimen is optimal for laboratory diagnosis of infectious diarrhea. A fresh stool sample is preferred for the identification of viral agents, protozoal agents, and C. difficile toxin. Molecular techniques generally are more sensitive and less dependent on the quality of the specimen than culture. Follow-up testing is not recommended in most patients following resolution of diarrhea. Collection and analysis of serial stool specimens using culture-dependent methods are recommended by local health authorities in certain situations.10
The Guidelines for the prevention and treatment of travelers’ diarrhea: a graded expert panel report14 indicates concerns regarding molecular testing and the detection of colonization rather than infection as well as the use of molecular diagnostic methods that do not allow for characterization of microbial resistance to antibiotics. Thus, the clinician may be in a position of ordering both culture-based and molecular tests to fully evaluate the patient’s illness and identify the best therapy. Further studies are needed to evaluate the utility of these assays in a clinical setting of the returning traveler with diarrhea.
Per the article, New molecular diagnostic tools in traveler’s diarrhea, the increase in worldwide tourism has led to an increase in imported diseases in patients returning to their home countries.13 Gastrointestinal problems remain the first complaint among travelers with health problems. Although most episodes of acute TD may resolve without the need for antimicrobial treatment and without identification of the causative agent(s), antibiotic treatment reduces the duration of diarrhea and related symptoms and also decreases time spent incapacitated. Bloody diarrhea and persistent diarrhea always require further investigation.13 Early diagnosis allows an appropriate antibiotic prescription, if needed, and helps to decrease unnecessary prescriptions. Antibiotic treatment can increase the risk of HUS in cases of STEC; therefore, rapid detection can have a significant impact on patient outcome through informing clinical decisions. An additional advantage of rapid diagnosis is the potential to improve public health response during outbreaks.
Molecular techniques for the routine diagnosis of diarrhea have been widely introduced in microbiological laboratories worldwide. An important advantage to multiplex testing is the automation of the laboratory workflow. Additional advantages include speed, less need for highly specialized technical personnel, minimal handling time, and reduced risk of contamination. Previous studies have shown the sensitivity of molecular assays is higher than culture and microscopy-based methods.5 Two of the main limitations of multiplex testing are an inability to distinguish between infection and colonization and the detection of an insignificant pathogen load that might not be involved with a patient’s clinical circumstances.6, 13 All microbiological results should be interpreted within a clinical context.13 While multiplex testing can detect lower levels of pathogenic organisms, false positive results can arise from detection of nonviable microorganisms, free DNA or RNA, or nonspecific amplifications.13 In addition, multiplex testing does not require culturing of live organisms, which can limit downstream organism typing studies required for outbreak assessments. Limitations in the detection of antibiotic resistance can be another pitfall in utilizing multiplex tests.13 Finally, the fixed composition of the majority of the molecular panels commercially available can limit the utility of these tests in less common disease presentations and contexts.13 There is a need for open multiplex molecular platforms where the consumer can combine different pathogens to be detected based on the local epidemiology, patient characteristics, and origin of the diarrhea.13
The health technology assessment of multiplex tests identifying gastrointestinal bacteria, viruses and parasites in patients with suspected infectious gastroenteritis16 concluded that in general, multiplex testing correctly identifies pathogens that are also identified by conventional testing. However, multiplex tests can also generate additional positive results not seen in conventional testing, which have uncertain clinical significance and potential to delay the patient’s ability to return to their job. Positive results could lead to unnecessary treatment when presenting symptoms would spontaneously resolve with watchful waiting. Negative results could lead to premature discharge from the hospital.
Multiple studies3,4,15-21 evaluated the FilmArray GIP panel. In most of the studies, the FilmArray specimens were submitted in Cary-Blair enteric transport medium and a few studies used frozen specimens. The studies used several different versions of the test, including a version for investigational use only, version 1.7, and version 2.0. The FilmArray GIP panel consists of automated nucleic acid extraction, reverse transcription, amplification, and analysis with results available in approximately one hour. The test detects 22 agents, including bacteria, viruses, and parasites. The studies compared the results of the multiplex panel with conventional laboratory techniques.
In all of the studies,3,4,15-21 the FilmArray GIP panel was positive for more pathogens than the conventional laboratory techniques and detected multiple pathogens in a single specimen at a higher rate than the conventional laboratory techniques. Limitations to the studies included a low number of positive specimens for some of the test targets, a lack of differentiation between live and dead organisms in FilmArray results, detection by FilmArray of organisms at non-pathogenic levels, unclear significance to clinical implications of detected coinfections, lack of clear distinction between detection of symptom-associated organisms versus asymptomatic shedding of organisms (colonization), and inability to untangle the influences and biases inherent to differences in geographic locations, seasons of sampling, and varied patient populations.
In the Buss et al3 clinical trial, the FilmArray GIP panel had 100% sensitivity/positive agreement (PPA) for 12 of the 22 targets and had greater than or equal to 94.5% sensitivity/PPA for an additional seven of 22 targets. It was not possible to assess the sensitivity/PPA for Vibrio spp., Vibrio cholerae, and Entamoeba histolytica, as the organisms were not detected by comparator methods, or at all in the case of Entamoeba histolytica. The specificity/negative percent agreement (NPA) was greater than or equal to 97.1% for all targets. In the Piralla et al17 trial, the PPA was 87.5% and the NPA was 77.1%.
In the Beal et al15 trial, researchers assessed the impact of the FilmArray GIP panel. The study found patients who had multiplex panel testing had fewer additional stool tests, fewer imaging studies, and fewer days on antibiotics. The overall length of stay was decreased by 0.5 days. Limitations of the study included absence of confirmatory testing when the results of multiplex panel tests and the conventional tests did not agree and the use of a historical cohort of patients for the control group as opposed to a contemporary group of patients alongside the test group.
In the Cybulski et al19 trial, the clinical impact of the FilmArray GIP panel was assessed. The investigators found patients with infections detected by the multiplex panel test were more likely to receive targeted rather than empirical therapy. Reduced time to treatment and an impact on antibiotic prescribing were also observed in STEC infections. Patients infected with pathogens identified both by culture and the multiplex panel test (concordant) tended to have greater symptom severity than those patients with positive multiplex panel tests and negative cultures (discordant). The investigators stated these observations were consistent with a higher organism burden required for positive cultures and the relationship of disease severity to the level of organism burden. Patients with coinfections were generally younger and were more likely to have recent international travel. Limitations to the study included an insufficient size of certain subgroups to draw definitive conclusions, a retrospective medical record review methodology that was subject to confounding factors and biases, and restriction of the study to only two hospitals within a single healthcare system.
The Khare et al8 trial compared two commercial multiplex panels for detection of gastrointestinal pathogens. Conventional laboratory techniques, testing on the FilmArray platform, and testing on the MagPix platform (xTag gastrointestinal pathogens panel) were performed on each specimen. The investigators followed the manufacturer’s instructions for the FilmArray platform; however, they modified the manufacturer’s FDA cleared protocol for the xTag panel by using Cary-Blair stool instead of raw stool. The investigators found that the majority of the targets represented on the panels showed high sensitivity and specificity; however, there were analytes on both panels that showed poorer performance. The FilmArray Aeromonas assay had low sensitivity and the xTag assay for Yersinia enterocolitica had low sensitivity. The xTag assay for norovirus had low specificity. This study also found a high percentage of stool samples positive for two or more pathogens. Limitations of the study included: different volumes of stool tested by the FilmArray assay (200 microliters) and the xTag assay (100 microliters), more dilute Cary-Blair specimens performed in the xTag sample rather than the FDA cleared raw stool, a lack of available clinical information and treatment decisions to compare with the test results, a relatively low number of positive results for some targeted organisms and an absence of positive results for other targets, and the lack of confirmation for some pathogens identified in the multiplex assays due to inadequate specimen remaining for confirmatory testing and/or unavailability of confirmatory testing.
Multiple studies6,9,22 compared Verigene, FilmArray, and xTag assays. Huang et al22 found that each of the assays demonstrated good clinical performance in their patient population. The Verigene and FilmArray assays provided rapid, on-demand testing for individual specimens in a moderate complexity environment. The xTag assay was higher complexity but allowed greater throughput in a single batch.
The Ramanan et al6 study found that in addition to broad coverage and the ability to identify a higher rate of coinfections, the multiplex panels offered reduced turnaround time. The authors supported that clinical laboratories need to be actively involved in the development of test utilization strategies focusing on the use and interpretation of the results of the tests as each of the panels had unique advantages and limitations. Conventional methods are still needed to detect pathogens not covered by the panels and provide antimicrobial susceptibility information.
The Binnicker9 study noted challenges associated with interpretation of the results of the multiplex panels. First, the multiplex panels target microbial nucleic acid and cannot distinguish between viable, replicating organisms responsible for disease and nonviable pathogens or remnant nucleic acid. The second challenge identified was the use of multiplex panels may increase the detection of C. difficile and potentially impact isolation and treatment decisions. The third challenge relates to how health care providers will use and interpret the large amount of data that will be made available with the broad implementation of the multiplex panels. In the conclusions, the author found each of the multiplex panels had unique advantages and limitations that a clinical microbiology lab must consider. Although the panels increased both the positivity rate and the number of coinfections detected, cultures will continue to be needed for antibiotic susceptibility and epidemiologic investigations. Future studies need to address clinically related issues, including the impact of the panels on antibiotic use and the influence the results have on management decisions and patient outcomes. In the author’s opinion, reserving the highly multiplexed panels to immunocompromised patients, the critically ill, or patients with prolonged diarrhea may improve test utilization and reduce inappropriate antibiotic use.
Summary of Evidence (Post Comment Period)
In the Axelrad et al 23 retrospective study, resource utilization was compared between patients receiving conventional stool testing (stool cultures with or without an ova and parasites exam or enzyme immunoassay for viral pathogens) and patients receiving GI panel testing (FilmArray GI pathogen panel). Both populations were from the same site (New York Presbyterian-Columbia University Medical Center), but were tested during two different time frames: conventional testing between 2012-2015 and GI panel testing between 2015-2017. Both outpatient and inpatient populations were included. Outcome metrics included the post-test likelihood to undergo endoscopy, to receive abdominal radiology, to be prescribed an antibiotic, to require a longer hospital stay, and to require a follow-up emergency department visit. Overall, the absolute risk of receiving endoscopy, abdominal radiology, and antibiotics was found to be lower in the GI panel testing population. This finding was thought to be associated with a higher rate of positive test results in the GI panel population. The researchers disclosed several limitations to their study including incomplete information on patient presentation and outcome and lack of thorough investigation into the cause and effect of positive findings and the patient’s signs and symptoms. Additionally, it should be noted that no significant discussion differentiating results and outcomes between outpatients and inpatients was identified while reviewing this publication.
In the Chang et al 24 meta-analysis, two different Polymerase Chain Reaction (PCR) multiplex-based tests (Luminex xTAG GPP and FilmArray GI panel) were compared across 11 studies and conclusions were drawn regarding the accuracy of both tests as compared to gold standards and regarding the performance of these two tests as compared to each other. The researchers concluded that both tests were “highly accurate” with FilmArray demonstrating overall “higher sensitivity and post-test probability” than xTAG GPP. However, the researchers stated that the meta-analysis did not clarify how these results would “translate to a clinical setting.”
In the Machiels et al 25 study, patients at the Rodboud University Medical Center were evaluated using “routinely performed” PCR panels. In parallel, the same specimens were evaluated by the FilmArray GI panel, but these results were not used to guide patient care. Then, the patients’ clinical courses were analyzed retrospectively, and the researchers theorized alternate clinical outcomes for these patients had the FilmArray GI panel been used instead of the PCR panels. The FilmArray GI panel was predicted to improve patient care. This prediction was extrapolated through comparing the number of pathogens detected by and the turn-around times of each test. Of note, the researchers stated “In the outpatient population, we found no impact on either antibiotic therapy or the number of prevented diagnostic procedures (data not shown).”25
Contractor Advisory Committee (CAC) Evidentiary Summary
After review of literature, the CAC advisory panel, which met in April of 2019, discussed GIP panels utilizing NAATs. In general, the biggest difference between the GIP multiplex NAAT testing panels is the number of targets. Viruses can shed even in asymptomatic patients and an issue with the multiplex NAATs testing panels is false positive results. The CAC panel emphasized history and physical are important as well as severity and duration of the patient’s symptoms in determining which test should be performed. The CAC panel mentioned overuse is always a concern with new tests and recommended these tests not be done routinely or at first encounter in an emergency department stating there should be criteria set such as fever, diarrhea, or bloody stools. The CAC panel considered the guidelines and noted differences between the current ACG and IDSA guideline recommendations. The CAC panel discussed that the GIP multiplex NAAT testing panels are not to be ordered on the otherwise healthy patient stating these tests are more appropriate for the patient who is already ill, such as transplant patients or immunosuppressed patients, when the provider needs specific information quickly.