Guideline for the Prevention and Control of Norovirus Gastroenteritis Outbreaks in Healthcare Settings, 2011

Q2: What are the best methods to identify a norovirus outbreak in a healthcare setting?

Diagnostic methods – Fecal specimens

EIA/ELISA

Author, Yr (Ref)	Study Design Quality	Study Objective	Population and Setting N	Results				Comments	Ref ID_Data extracted by
Khamrin, P; 2008 120	Diagnostic Study   1,2	To evaluate the test characteristics of immunochromatography and ELISA (Denka) when compared with monoplex RT-PCR for detection of norovirus from stool specimens.	Infants and children with acute gastroenteritis in Japan   503 fecal specimens	Test characteristics of immunochromatography and ELISA Immunochromatography vs. RT-PCR TP – 90 TN – 375 False positive (FP) – 14 False negative (FN) – 24 Sensitivity – 78.9% Specificity – 96.4% PPV – 86.5% NPV – 94.0% Accuracy – 92.4%   ELISA vs. RT-PCR TP – 103 TN – 375 FP – 14 FN – 11 Sensitivity – 90.4% Specificity – 96.4% PPV – 88.0% NPV – 97.2% Accuracy – 95.0%   Accuracy of norovirus genotype detection All results listed as positives detected/true positives Immunochromatography vs. RT-PCR GI/1 – 1/2 GII/3 – 13/14 GII/4 – 75/95 GII/6 – 1/3   ELISA vs. RT-PCR GI/1 – 2/2 GII/3 – 12/14 GII/4 – 86/95 GII/6 – 3/3				Immunochromatography takes 20 min. ELISA takes 4 hrs.   Power and sample size not reported   Prevalence not reported	2351_RA
Wiechers, C; 2008 121	Descriptive study   1,2,3	To describe a cluster of positive IDEIA cases which were unable to be confirmed using RT-PCR or EM.	Infants in a level III neonatal intensive care unit in Germany during November 2003.   43 infants screened.   163 stool samples obtained.	# positive/# tested samples IDEIA: 46/163 samples from 22/43 infants were positive. RT-PCR: 0/11 samples with enough volume were positive. EM: 0/11 samples were positive.   Variables associated with IDEIA positive samples Stools with and without blood: 11/46 vs. 1/117; p<0.001 Age of patients with IDEIA positive vs. negative samples: median 34.9 weeks (range 28.6-40.9) vs. 36.6 weeks (range 29.4-66.9); p<0.001.				RT-PCR (QIAGEN, Hilden, Germany). IDEIA NLV kit (DakoCytomation Ltd., Ely, UK).	5118_IL
Castriciano S, 2007 122	Diagnostic Study   1,2,3	To compare RIDASCREEN norovirus EIA to IDEIA NLV GI/GII	66 positive and 162 negative stool samples   228 total samples	Test characteristics: Test – Positive (% sensitivity; CI) vs. Negative (% specificity; CI) RT-PCR: 65 (98.5; 91.9-99.7) vs. 162 (100; 97.7-100) RIDASCREEN: 53 (80.3; 69.2-88.1) vs. 162 (100; 97.7-100) IDEIA-NLV: 40 (60.6; 48.5-71.5) vs. 162 (100; 97.7-100) EM: 24 (36.4; 25.8-48.4) vs. 157 (96.9; 93.0-98.7)				Used stools that had previously been screened by EM and stored at -70 C. Re-tested using RT-PCR.	143_IL
Gray, JJ; 2007 118	Diagnostic study   2,3	To determine test characteristics for IDEIA and RIDA-SCREEN.	Stool samples from patients with symptoms of gastroenteritis collected during the 2004-2005 and 2005-2006 norovirus seasons and evaluated in this European multicenter study.   2,254 samples from 273 outbreaks.   274 samples collected in sporadic cases.   144 samples had other enteric pathogens identified.	Test characteristics IDEIA: Sensitivity 58.93% (95% CI 56.12-61.68%) Specificity 93.91% (95% CI 92.23-95.25%) PPV 92.30% NPV 64.90%   RIDA-SCREEN: Sensitivity 43.81% (95% CI 41.01-46.65%) Specificity 96.27% (95% CI 95.00-97.38%) PPV 93.70% NPV 58.20%   Sensitivity for differing number of samples tested The sensitivity for outbreak diagnosis improved when ≥6 samples tested. IDEIA: 3 vs. 6 samples tested (z=±3.191; p=0.0014) RIDA-SCREEN: 3 vs. 6 samples tested (z=±3.828; p=0.0001)   Range of norovirus genotypes detected All samples: Genotype - IDEIA vs. RIDASCREEN No [(%) samples genotype detected (95% CI)]; p value GI-1 – 4 [80.00% (37.55-96.36%)] vs. 3 [60.00% (23.07-88.24%)]; 0.49 GI-2 – 11 [84.62% (57.77-95.67%)] vs. 2 [15.38% (4.33-42.23%)]; 0.0002 GI-3 – 12 [42.86% (26.51-60.93%)] vs. 9 [32.14% (17.93-50.66%)]; 0.4 GI-4 – 2 [100.00% (34.24-100.00%)] vs. 0 [0.00% (0.00-65.76%)]; 0.3 GI-5 – 3 [37.50% (13.68-69.43%)] vs. 0 [0.00% (0.00-32.44%)]; 0.2 GI-6 – 5 [71.43% (35.89-91.78%)] vs. 0 [0.00% (0.00-35.43%)]; 0.02 GI-7 – 0 [0.00% (0.00-79.35%)] vs. 0 [0.00% (0.00-79.35%)]; >0.5 GII-1 – 7 [87.50% (52.91-97.76%)] vs. 0 [0.00% (0.00-32.44%)]; 0.0024 GII-2 – 8 [50.00% (28.00-72.00%)] vs. 4 [25.00% (10.18-49.50%)]; 0.2 GII-3 – 30 [57.69% (44.19-70.13%)] vs. 11 [21.15% (12.24-34.03%)]; 0.0003 GII-4 – 203 [67.44% (61.96-72.49%)] vs. 186 [61.79% (56.19-67.10%)]; 0.17 GII-5 – 2 [33.33% (9.68-70.00%)] vs. 3 [16.67% (3.01-56.35%)]; >0.5 GII-6 – 2 [22.22% (6.32-54.74%)] vs. 0 [0.00% (0.00-29.91%)]; 0.4 GII-7 – 20 [68.97% (50.77-82.72%)] vs. 5 [17.24% (7.6-34.55%)]; 0.002 GII-8 – 0 [0.00% (0.00-79.35%)] vs. 0 [0.00% (0.00-79.35%)]; >0.5 GIV-1 – 0 [0.00% (0.00-48.99%)] vs. 0 [0.00% (0.00-48.99%)]; >0.5 rGII – 10 [52.63% (31.71-72.67%)] vs. 2 [10.53% (2.94-31.39%)]; 0.01   IDEIA showed reactivity to a broader range of genotypes than the RIDASCREEN norovirus assay, which showed genotype-dependent sensitivities.				IDEIA norovirus (Oxoid; Thermo Fisher Scientific, Ely, UK). RIDASCREEN norovirus (R-Biopharm, Darmstadt, Germany) RT-PCR was the reference standard.	053_IL
Wilhelmi de Cal, I; 2007 123	Diagnostic study   2,3	To evaluate IDEIA and Ridascreen compared to RT-PCR for norovirus antigen detection.	The study included stool samples from children <5 years of age with acute gastroenteritis who were admitted to a hospital in Spain between October 1, 2002 and April 1, 2004.   Stools collected 24-48 hrs after admission with a diagnosis of acute gastroenteritis   117 samples that were negative for bacterial pathogens, rotaviruses, adenoviruses, and astroviruses were tested for Caliciviridae by RTPCR, IDEIA, and Ridascreen.	Samples positive for norovirus 39 samples positive by RT-PCR. Concordant results with 3 methods in 77 (65.8%) samples. Discordant results with 3 methods in 40 (34.2%) samples. 18/39 samples underwent genotyping and sequence analysis: 1 had Sapovirus and 17 were norovirus genogroup II.   Test characteristics IDEIA: Sensitivity 76.9% Specificity 85.9% PPV 73.2% NPV 88.2% Agreement 82.9% Kappa index 0.6203   Ridascreen: Sensitivity 59% Specificity 73.1% PPV 52.3% NPV 78.1% Agreement 68.4% Kappa index 0.3103				IDEIA NVL assay (DakoCytomation, Ely, UK). Ridascreen NLV (R-BioPharm, Darmstadt, Germany). RT-PCR assay (One-Step RT-PCR Kit, QIAGEN, Valencia, CA, USA).	144_IL
De Bruin, E; 2006 124	Diagnostic study   2,3	To evaluate IDEIA and Ridascreen EIAs compared to RT-PCR for the diagnosis of acute gastroenteritis outbreaks.	Two panels of stool samples collected from Dutch gastroenteritis surveillance (1999 2003).   Panel 1: 158 fecal samples from 23 outbreaks, including confirmed Rotavirus and Astrovirus outbreaks that had been tested for norovirus by RT-PCR in 2002 and 2003.   Panel 2: 19 samples positive for norovirus by RT-PCR: 6 samples of 5 different genogroup I strains, 12 samples of 6 genogroup II strains, and 1 genogroup IV strain. These stool samples were collected from Dutch gastroenteritis surveillance from 1999 to 2002.	Agreement between ELISAs and RT-PCR Positive in all tests – 10/158 (6%) Negative in all tests – 71/158 (45%) Discrepant results – 77/158 (49%)   Detection of norovirus Samples with ELISA kits 1. ELISA (Dako kit) vs. RT-PCR (All samples) TP – 28 TN - 81 FN – 46 FP – 3 Sensitivity – 37.8% Specificity – 96.4% PPV – 90.3% NPV – 63.8%   2. ELISA (Dako kit) vs. RT-PCR (norovirus positive outbreaks) Criterion A – Two or more norovirus positive samples per outbreak to identify the causative agent TP – 30 TN – 40 FP – 1 FN – 43   Criterion B – 50% or more norovirus positive samples per outbreak to identify the causative agent TP – 24 TN – 63 FP – 7 FN – 38   3. ELISA (Ridascreen kit) vs. RT-PCR TP – 27 TN - 74 FN – 47 FP – 10 Sensitivity – 36.5% Specificity – 88.1% PPV – 73.0% NPV – 61.2%   4. ELISA (Ridascreen kit) vs. RT-PCR (norovirus positive outbreaks) Criterion A – Two or more norovirus positive samples per outbreak to identify the causative agent TP – 35 TN – 39 FP – 2 FN – 38   Criterion B – 50% or more norovirus positive samples per outbreak to identify the causative agent TP – 29 TN – 62 FP – 8 FN – 33   Detection of norovirus outbreaks with ELISA kits 1. ELISA (Dako kit) vs. RT-PCR Criterion A – Two or more norovirus positive samples per outbreak to identify the causative agent TP – 8 TN – 8 FP – 0 FN – 7   Criterion B – 50% or more norovirus positive samples per outbreak to identify the causative agent TP – 5 TN – 11 FP – 0 FN – 7   2. ELISA (Ridascreen kit) vs. RT-PCR Criterion A – Two or more norovirus positive samples per outbreak to identify the causative agent TP – 9 TN – 8 FP – 0 FN – 6   Criterion B – 50% or more norovirus positive samples per outbreak to identify the causative agent TP – 4 TN – 11 FP – 0 FN – 8   RIDASCREEN not able to discriminate between groups 17% of PCR-identified Genogroup I 58% of PCR-identified Genogroup II 0% of PCR-identified by Genogroup IV   74/158 samples confirmed NLV via PCR and Southern Blot Of these, 28/74 confirmed with Dako and 27/74 with RIDAscreen   84/158 samples were negative by PCR 3/84 negative by PCR were positive using Dako 10/84 negative by PCR were positive using RIDAscreen Dako: 96% specificity Ridascreen: 88% specificity				RT-PCR protocol followed by Southern blot hybridization was the reference standard.   IDEIA (DakoCytomation Ltd., Ely, UK). Ridascreen (R-biopharm AG, Darmstadt, Germany).   Prevalence not reported	238_IL
Okitsu-Negishi, S; 2006 125	Diagnostic study   2,3	To evaluate the RIDASCREEN norovirus ELISA kit compared to RTPCR.	503 stool samples collected from infants and children with acute sporadic gastroenteritis who visited 6 pediatric clinics in Japan from July 2004 to March 2005.	Test characteristics for RIDASCREEN Sensitivity - 76.3% Specificity - 94.9%. PPV - 81.3% NPV – 93.2% 90.7% agreement   FP - 20 TP -87 FN - 27 TN - 369   Sensitivity by norovirus genotype All results – # positive/# tested (%) GI/1 – 1/2 (50%) GII/3 – 3/13 (23.1%) GII/4 – 82/96 (85.4%) GII/6 – 1/3 (33.3%)				RT-PCR was the reference standard. RIDASCREEN (R-Biopharm AG, Darmstadt, Germany).   Power and sample size not reported.   Prevalence not reported	228_IL
Burton- MacLeod, JA: 2004 126	Diagnostic study   2,3	To assess two enzyme linked immunosorbent assay kits, SRSV (II)-AD and IDEIA, compared to RTPCR.	104 stool samples with norovirus: 4 genogroup I subgroups and 10 genogroups II subgroups from 35 outbreaks that occurred in the US June 1999-2002.   33 samples with other enteric viruses from children <5 years of age with diarrhea.   SRSV (II)-AD also tested with 6 Sapovirus positive samples from patients in an outbreak.	Test characteristics S RSV (II)-AD: Sensitivity 80% Specificity 69% 77% agreement Sensitivities > 70% for 10/14 subgroups Cross-reacted with samples containing norovirus GI and GII subgroups; as well as samples with human Sapovirus. Detected 59% of the GII antigens in the GI wells and 63% of the GI antigens in the GII wells.   IDEIA: Sensitivity 39% Specificity 100% 54% agreement Sensitivities >70% for 3/14 subgroups. GII/2, GII/5, GII/6, and GII/n may not be detected by IDEIA. Discriminated between norovirus GI and GII antigens. Detected no GII antigens in the GI wells and only 7% of GI antigens in the GII wells.				SRSV (II)-AD (Denka Seiken Co. Ltd., Tokyo, Japan). IDEIA NLV (DakoCytomation Ltd., Ely, UK). RT-PCR was the reference standard.   Power and sample size not reported.	660_IL
Christen, A; 2003 127	Diagnostic study   2,3, 4	To evaluate IDEIA compared to RT-PCR in detecting norovirus.	39 stool samples from a prior case-control study conducted in Switzerland. 24 additional samples previously PCR tested by a German Laboratory.	Swiss samples TP – 9 TN – 15 FN - 12 FP - 3   IDEIA Test characteristics Sensitivity 0.43 Specificity 0.83 PPV 0.75 NPV 0.56   Relative trueness 0.62 False positive 0.17 False negative 0.57 Concordance index Kappa 0.25   German samples TP – 6 TN – 11 FN - 7 FP – 0   IDEIA Test characteristics Sensitivity 0.46 Specificity 1.00 PPV 1.00 NPV 0.61   Relative trueness 0.71 False positive 0.00 False negative 0.54 Concordance index Kappa 0.44				IDEIA NLV ELISA (Dako-Cytomation, Ely, UK). RT-PCR was the reference standard. Power and sample size not reported. Prevalence not reported Differences in sensitivities may have resulted from differences in storage of samples (4°C for <3 days versus -20°C for long term storage as recommended by the manufacturer). Some samples had been stored for many weeks at 4°C.	4519_IL
Gunson, R; 2003; 128	Diagnostic study   1	To compare a real-time polymerase chain reaction (PCR) and a newly developed EIA for the detection of norovirus. Negative or discrepant PCR results were investigated using EM and a different, not real time PCR.	Stool samples were collected from outbreaks and sporadic cases/unidentified outbreaks, no timeframe specified   70 stool samples	Positive samples detected 1. PCR Overall – 26 Among sporadic cases – 5   2. EIA Overall – 10 Among sporadic cases – 3   All PCR samples could be confirmed using the second PCR. The EIA detected two positive samples that were negative by the PCR. Neither of these samples could be confirmed using the second PCR or EM.				Power and sample size not reported.   Prevalence not reported	757_RA
				Real-time PCR	EIA	Positive	Negative
					Positive	8	18
					Negative	2	42
				Test characteristics (%) Sensitivity – 30.8 Specificity – 95.5 PPV – 80.0 NPV – 70.0
Rabenau, H; 2003 17	Diagnostic study   1, 2	To compare the sensitivity and specificity of: 1. ELISA when compared with a) TEM and PCR or b) PCR only 2. TEM when compared with a) ELISA and PCR or b) PCR only	Inhabitants and employees of homes for the elderly (in Frankfurt, Germany) aged 20 to >60 years; 73% females, 42% > 60 yrs.   244 stool samples from 227 patients	Test characteristics (%) for ELISA When compared with TEM and PCR Sensitivity – 50.0 Specificity – 96.2 PPV – 68.0 NPV – 92.2 (True Positive[TP] – 17; True Negative[TN] – 202; FP – 8; FN – 17) When compared with PCR only Sensitivity – 31.3 Specificity – 94.9 PPV – 60.0 NPV – 84.9 (TP – 15; TN – 186; FP – 10; FN – 33)   Test characteristics (%) for TEM When compared with ELISA and PCR Sensitivity – 88.2 Specificity – 99.0 PPV – 93.8 NPV – 98.1 (TP – 30; TN – 208; FP – 2; FN – 4) When compared with PCR only Sensitivity – 58.3 Specificity – 98.0 PPV – 87.5 NPV – 90.6 (TP – 28; TN – 192; FP – 4; FN – 20)   Test characteristics (%) for PCR When compared with ELISA and TEM Sensitivity – 94.1 Specificity – 92.4 PPV – 66.7 NPV – 99.0 (TP – 32; TN – 194; FP – 16; FN – 2)				Power and sample size not reported.   Prevalence not reported	801_RA
Richards, A; 2003 119	Diagnostic Study   1,2	To determine the test characteristics of ELISA and EM in detecting norwalk-like virus (NLV) infection when compared with PCR	Fecal samples collected from patients involved in outbreaks of gastroenteritis in the UK   531 fecal samples	Test characteristics (%) of ELISA vs. PCR Sensitivity – 55.5(51.1-60.0) Specificity – 98.3(97.1-99.9) PPV – 95.0(CI not reported) NPV – 76.9 (CI not reported)   Test characteristics (%) of EM vs. PCR Sensitivity – 23.9(19.5-28.1) Specificity – 99.2(98.3-100) PPV – 93.9(CI not reported) NPV – 70.7(CI not reported)   Identification of NLV as the cause of an outbreak (% of outbreaks) When the causative agent was defined by ≥ 2 positive samples EM – 7.2 ELISA – 18.6 PCR – 41.5 When the causative agent was defined by ≥ 1 positive samples EM – 19.6 ELISA – 47.8 PCR – 62.8   Sensitivity; Specificity of ELISA based on number of samples collected 2 samples – 52.9; 100 ≥4 – 69.2; 100 ≥6 – 71.4; 100   Other results Agreement between ELISA and PCR – 81.8% (Kappa = 0.57) Sensitivity of ELISA was significantly increased when compared with EM (P<0.01)				Power and sample size not reported.   Prevalence not reported	848_RA

Next Table