- Email: [email protected]
Clinical impact of timely reporting of IgM for West Nile Virus
Journal of Clinical Virology 34 (2005) 122–124
Clinical impact of timely reporting of IgM for West Nile Virus J. Barenfanger ∗ , C. Drake, J. Lawhorn, J. O’Brien, T. Mueller Laboratory Medicine, Memorial Medical Center, 701N First Street, Springfield, IL 62781, USA Received 15 October 2004; received in revised form 24 January 2005; accepted 2 February 2005
Abstract Background: In 2002 diagnostic testing for West Nile Virus (WNV) infection was not ideal because commercial products were unavailable. Although testing was performed without charge by public health (PH), testing is not intended to be used for patient management and generally results are not available in a timely or reliable manner. Objectives: This study was conducted to determine the clinical impact of having WNV serology available by a diagnostic laboratory within a few days of sample collection. Study design: An IgM immunoassay (FT, from Focus Technologies) was performed August 13, 2003 to October 31, 2003 Monday, Wednesday and Friday in our hospital’s virology laboratory with split sample testing at PH. All results (positive and negative) were called to physicians. Results and conclusions: Turn-around time for the FT assay was within 0–3 days; for PH, it was ∼8 days. Of 95 samples, 6 were positive and 89 negative. For the FT negative samples, all were concordant with PH, but four FT positive samples (three sera and one cerebrospinal fluid) were discrepant. After resolution of the discrepancies, the FT test had a sensitivity of 100% and a specificity of 99%. Physicians reacted positively to having results faster. Fifty-eight percent of the results had clinical impact, 22% had undetermined impact, 13% were patient requests, and 4% had no impact. Concluding, the FT assay is 100% sensitive and 99% specific and in 58% there was clinical impact for WNV reporting in a timely manner. © 2005 Elsevier B.V. All rights reserved. Keywords: Timeliness; Usefulness; West Nile Virus serology
1. Introduction In 2002 Illinois had 884 documented cases of West Nile Virus (WNV) infection with 66 deaths, more than anywhere in the U.S. that year (Makar and Stowell, 2004 and www.idph.il.us.gov). At that time, diagnostic testing for IgM activity against WNV was far from ideal because commercial products were not available. Although testing was performed without charge by the state public health laboratory (PH), this type of testing is not intended for patient management and generally results are not available in a timely or reliable manner as is the usual case with testing done by diagnostic laboratories. In addition to delayed and erratic turn-around times (often greater than several weeks in Abbreviations: WNV, West Nile Virus; PH, public health laboratory; FT, Focus Technology IgM serology for WNV ∗ Corresponding author. Tel.: +1 217 788 3000; fax: +1 217 788 5577. E-mail address: [email protected] (J. Barenfanger). 1386-6532/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jcv.2005.02.007
2002), appropriate health care workers were not informed and reports could be confusing (personal observation in 2002). This study was conducted during the summer of 2003 to determine the clinical impact of having the results for the IgM serology available within a few days after collection of sample and reported in a manner in which diagnostic laboratories usually do. This work was presented in part at 104th General Meetings at American Society for Microbiology, 2004, Abstr. C-050, Final Program Book, p. 51.
2. Materials and methods 2.1. Preparation work Approval from an Institutional Board Review (IRB) was obtained for this study. The IRB deemed that informed con-
J. Barenfanger et al. / Journal of Clinical Virology 34 (2005) 122–124
123
sent was not considered necessary. A newsletter sent to physicians said that all results (positive and negative) would be telephoned in order to determine the clinical impact of timely reporting. The newsletter stated that the sensitivity was 99% and that there was substantial cross reactivity with other Flaviviruses like those causing dengue fever and St. Louis encephalitis (25% and 40%, respectively), but since both these agents cause a viral encephalitis with similar management, this cross-reactivity may not be a real problem.
3. Results
2.2. IgM testing
The results from our FT testing and the state laboratory testing were concordant for all the FT negative samples, but four of the FT positive samples (three sera and one CSF) were discrepant with PH.
The IgM immunoassay (FT, Focus Technologies, Cypress, CA) was performed on serum or cerebrospinal fluid (CSF) from August 13, 2003 to October 31, 2003 Monday, Wednesday and Friday with split sample testing at PH. The procedure for the IgM capture enzyme-linked immunoabsorbent assay for WNV followed the procedure provided by the manufacturer (Product Code EL0300M, Focus Technologies, Cypress, CA). This procedure is detailed in another study (Hogrefe et al., 2004). Briefly, dilutions of the sample were made, four incubations and several washings were performed, then samples were read spectrophotomically at λ = 450 nm. A formula was used to convert the reading into an index value. A result of <0.9 index value was negative. The test took ∼4 h in total to perform, with ∼1 h (or less) of hands-on time. Split sample testing was done by the state public health laboratory. Samples were sent to the state laboratory Monday–Thursday. Turn-around time was calculated from the day the specimen was obtained from the patient to the time the result was available in the medical record. The same type of test (IgM capture immunoassay) was used, but the reagents the state laboratory used came in part from the Centers of Disease Control and in part from Focus Technologies. 2.3. Heterophile test absorbance Samples initially IgM positive could be tested again in the same IgM test as described above using both a “control” well and an “antigen” well to evaluate for heterophile activity in the sample. Briefly, diluent was added to one patient well (control well) and WNV antigen was added to the second patient well (antigen well). The standard procedure for the IgM immunoassay (described above) was followed. At the completion of the entire procedure, the absorbance of the control well was subtracted from the absorbance of the antigen well to obtain the net absorbance. 2.4. IgG testing The IgG test for WNV was performed by Focus Technologies, Cypress, CA.
3.1. Turn-around time Turn-around time for the FT assay performed at this institution was within 0–3 days; for PH, it was ∼8 days. Eightyone sera and 14 CSF were tested. FT testing had 5 sera and 1 CSF positive; 76 sera and 13 CSF, negative. 3.2. Comparison testing
3.3. Resolution of discrepancies The three sera which tested positive by the FT kit but negative by PH were retested to detect the presence of heterophile antibodies using the absorbance technique. Two samples remained positive while one sample became negative, indicating that the heterophile antibody was the cause of one FT positive test. The other three discrepant samples were PH false negatives. We know this because (a) for the CSF discrepancy, the patient’s serum was positive by both FT and PH and (b) the other two sera had positive confirmatory IgM tests and convalescent sera which converted from IgG negative to positive. In addition, all of the three latter discrepant samples had clinical history highly consistent with WNV. After resolution of the discrepancies, sensitivity of the FT test was 100% (5/5) and specificity 99% (89/90). 3.4. Clinical impact Generally physicians were contacted by telephone within 0–2 days of the test to communicate the results and to obtain their assessment of the clinical impact of the information. The vast majority of the time, physicians were contacted directly on the same day the test was performed. Alternatively, nurses were contacted if the physicians were not available. Fifty-five (58%) had clinical impact, 21 (22%) had undetermined impact, 12 (13%) were patient requests, 4 (4%) had no impact; 1 was done because the serum came from a potential donor of blood/tissue and on 2, the physician said the test was unnecessary but the test was ordered under the physician’s name (Table 1). Physicians generally were most appreciative of having the results in a more timely manner. Clinical impact for patients who tested positive included such reasons as (1) a patient was a candidate for a treatment study for WNV and (2) anti-viral therapy for Herpes simplex could be discontinued. Clinical impact for patients who tested negative was mainly that the information helped rule out WNV and set the workup in a different direction. In one case, the
124
J. Barenfanger et al. / Journal of Clinical Virology 34 (2005) 122–124
Table 1 Clinical impact of timely reporting of WNV serology Significance of test
Number of specimens (%)
Positive clinical impact Patient request Undetermined No impact Unnecessary Donor Total
55 (58) 12 (13) 21 (22) 4 (4) 2 (2) 1 (1) 95 (100)
patient was a candidate for an anti-dementia drug since WNV had been excluded.
4. Discussion Although there are many exceptions, most public health labs (by virtue of their construct as a tertiary reference laboratory) generally will not have the responsiveness or turnaround time of an onsite clinical laboratory. (The exceptions include such as expedited testing of rabies, hantavirus and SARS. In fact, in California WNV testing is often completed in a few days by a network of state and local public health laboratories and results are immediately relayed to clinicians.) However, it is possible that public health laboratories are used for tests intended to help manage patients (which is not the usual function of such public health labs). This study supports the concept that a rapid turn-around time for this test is clinically useful. Currently, two FDA-approved IgM Capture ELISA diagnostic kits are available to clinical laboratories for the serological diagnosis of West Nile Virus—one from Focus Technologies used in this study and another from PanBio Limited (Windsor, Australia). Our findings of the performance of the test are consistent with other studies. Moore et al. (2003) found a sensitivity of 99.6% and Prince et al. (2004) found a sensitivity and specificity of 100%. All of the FT negative samples had 100% concordance with PH test results. Only one of the four discrepant FT positive samples was a FT false positive (the patient had heterophile antibody which may cause a false positive for WNV). Heterophile antibodies present in human sera may react with proteins, typically immunoglobulin, from non-human species such as goats, horses, rabbits and mice. Although estimates of the prevalence of heterophile antibodies in normal sera vary from 0.5% to as high 20%, most think the prevalence is <2%. In rare circumstances, heterophile antibodies can be a cause of false positive results in serology tests. This is due to the fact that many laboratory tests utilize goat, rabbit or mouse sera as components of the reagents. If the absorbance test had been performed after the test was first determined to be positive (as the manufacturer recommends),
then the heterophile antibody would have been detected and the specimen would not have been determined to be positive. The other three discrepant FT positive samples were PH false negatives. Although Illinois was the hardest hit for WNV in 2002, it had substantially fewer cases at the time of this study, with only 54 documented cases and 1 death compared to 884 cases and 66 deaths the previous year (www.idph.il.us.gov). However, physicians were still most appreciative of notification of the test result, even when the vast majority of the results were negative. Just over half physicians responded in a manner suggesting that there was clinical impact for the information. This implies that it would be clinically useful to offer this test in a timely manner and that clinical diagnostic laboratories may not want to rely on testing from the public health laboratory in Illinois. Since rapid turn-around time is not a priority for most public health laboratories, this issue mostly likely pertains in other states as well. In one instance we described, the positive result of WNV enabled a physician to discontinue acyclovir. It is quite possible that patients with unknown encephalitidies are on anti-bacterial, anti-TB and anti-fungal medications as well. Therefore, a positive test result may impact management beyond just the use of acyclovir. Additionally, finding an etiologic agent is often very important to family members. In conclusion, the sensitivity of the FT test was 100% and specificity 99%, and this test was determined to have a clinical impact in 58% of the cases. As a result, this diagnostic laboratory offered WNV testing for three times a week during the peak mosquito season in 2004.
Acknowledgement This study was supported in part by Focus Technologies, Cypress, CA.
References Hogrefe WR, Moore R, Lape-Nixon M, Wagner M, Prince HE. The performance of West Nile Virus recombinant (preM/E)-based IgG and IgM ELISA for the detection of West Nile Virus and other flavivirus antibodies. J Clin Microbiol 2004;42:4641–8. Makar RS, Stowell CP. West Nile Virus: emerging infection in transfusion medicine. Clin Microbiol Newsl 2004;26:65–70. Moore RJ, Prince HE, Lape-Nixon M, Hogrefe WR. Comparison of the Focus Technologies West Nile Virus IgG and IgM kit results to public health serology results. In: Proceedings of the 52nd Annual Meeting of the American Society of Tropical Medicine and Hygiene; 2003. Prince H, Lape-Nixon MR, Moore RJ, Hogrefe WR. Utility of the focus technologies West Nile Virus IgM capture enzyme-linked immunosorbent assay for testing cerebrospinal fluid. J Clin Microbiol 2004;42:12–5 May 11, 2004, www.idph.il.us.gov.
Clinical impact of timely reporting of IgM for West Nile Virus J. Barenfanger ∗ , C. Drake, J. Lawhorn, J. O’Brien, T. Mueller Laboratory Medicine, Memorial Medical Center, 701N First Street, Springfield, IL 62781, USA Received 15 October 2004; received in revised form 24 January 2005; accepted 2 February 2005
Abstract Background: In 2002 diagnostic testing for West Nile Virus (WNV) infection was not ideal because commercial products were unavailable. Although testing was performed without charge by public health (PH), testing is not intended to be used for patient management and generally results are not available in a timely or reliable manner. Objectives: This study was conducted to determine the clinical impact of having WNV serology available by a diagnostic laboratory within a few days of sample collection. Study design: An IgM immunoassay (FT, from Focus Technologies) was performed August 13, 2003 to October 31, 2003 Monday, Wednesday and Friday in our hospital’s virology laboratory with split sample testing at PH. All results (positive and negative) were called to physicians. Results and conclusions: Turn-around time for the FT assay was within 0–3 days; for PH, it was ∼8 days. Of 95 samples, 6 were positive and 89 negative. For the FT negative samples, all were concordant with PH, but four FT positive samples (three sera and one cerebrospinal fluid) were discrepant. After resolution of the discrepancies, the FT test had a sensitivity of 100% and a specificity of 99%. Physicians reacted positively to having results faster. Fifty-eight percent of the results had clinical impact, 22% had undetermined impact, 13% were patient requests, and 4% had no impact. Concluding, the FT assay is 100% sensitive and 99% specific and in 58% there was clinical impact for WNV reporting in a timely manner. © 2005 Elsevier B.V. All rights reserved. Keywords: Timeliness; Usefulness; West Nile Virus serology
1. Introduction In 2002 Illinois had 884 documented cases of West Nile Virus (WNV) infection with 66 deaths, more than anywhere in the U.S. that year (Makar and Stowell, 2004 and www.idph.il.us.gov). At that time, diagnostic testing for IgM activity against WNV was far from ideal because commercial products were not available. Although testing was performed without charge by the state public health laboratory (PH), this type of testing is not intended for patient management and generally results are not available in a timely or reliable manner as is the usual case with testing done by diagnostic laboratories. In addition to delayed and erratic turn-around times (often greater than several weeks in Abbreviations: WNV, West Nile Virus; PH, public health laboratory; FT, Focus Technology IgM serology for WNV ∗ Corresponding author. Tel.: +1 217 788 3000; fax: +1 217 788 5577. E-mail address: [email protected] (J. Barenfanger). 1386-6532/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jcv.2005.02.007
2002), appropriate health care workers were not informed and reports could be confusing (personal observation in 2002). This study was conducted during the summer of 2003 to determine the clinical impact of having the results for the IgM serology available within a few days after collection of sample and reported in a manner in which diagnostic laboratories usually do. This work was presented in part at 104th General Meetings at American Society for Microbiology, 2004, Abstr. C-050, Final Program Book, p. 51.
2. Materials and methods 2.1. Preparation work Approval from an Institutional Board Review (IRB) was obtained for this study. The IRB deemed that informed con-
J. Barenfanger et al. / Journal of Clinical Virology 34 (2005) 122–124
123
sent was not considered necessary. A newsletter sent to physicians said that all results (positive and negative) would be telephoned in order to determine the clinical impact of timely reporting. The newsletter stated that the sensitivity was 99% and that there was substantial cross reactivity with other Flaviviruses like those causing dengue fever and St. Louis encephalitis (25% and 40%, respectively), but since both these agents cause a viral encephalitis with similar management, this cross-reactivity may not be a real problem.
3. Results
2.2. IgM testing
The results from our FT testing and the state laboratory testing were concordant for all the FT negative samples, but four of the FT positive samples (three sera and one CSF) were discrepant with PH.
The IgM immunoassay (FT, Focus Technologies, Cypress, CA) was performed on serum or cerebrospinal fluid (CSF) from August 13, 2003 to October 31, 2003 Monday, Wednesday and Friday with split sample testing at PH. The procedure for the IgM capture enzyme-linked immunoabsorbent assay for WNV followed the procedure provided by the manufacturer (Product Code EL0300M, Focus Technologies, Cypress, CA). This procedure is detailed in another study (Hogrefe et al., 2004). Briefly, dilutions of the sample were made, four incubations and several washings were performed, then samples were read spectrophotomically at λ = 450 nm. A formula was used to convert the reading into an index value. A result of <0.9 index value was negative. The test took ∼4 h in total to perform, with ∼1 h (or less) of hands-on time. Split sample testing was done by the state public health laboratory. Samples were sent to the state laboratory Monday–Thursday. Turn-around time was calculated from the day the specimen was obtained from the patient to the time the result was available in the medical record. The same type of test (IgM capture immunoassay) was used, but the reagents the state laboratory used came in part from the Centers of Disease Control and in part from Focus Technologies. 2.3. Heterophile test absorbance Samples initially IgM positive could be tested again in the same IgM test as described above using both a “control” well and an “antigen” well to evaluate for heterophile activity in the sample. Briefly, diluent was added to one patient well (control well) and WNV antigen was added to the second patient well (antigen well). The standard procedure for the IgM immunoassay (described above) was followed. At the completion of the entire procedure, the absorbance of the control well was subtracted from the absorbance of the antigen well to obtain the net absorbance. 2.4. IgG testing The IgG test for WNV was performed by Focus Technologies, Cypress, CA.
3.1. Turn-around time Turn-around time for the FT assay performed at this institution was within 0–3 days; for PH, it was ∼8 days. Eightyone sera and 14 CSF were tested. FT testing had 5 sera and 1 CSF positive; 76 sera and 13 CSF, negative. 3.2. Comparison testing
3.3. Resolution of discrepancies The three sera which tested positive by the FT kit but negative by PH were retested to detect the presence of heterophile antibodies using the absorbance technique. Two samples remained positive while one sample became negative, indicating that the heterophile antibody was the cause of one FT positive test. The other three discrepant samples were PH false negatives. We know this because (a) for the CSF discrepancy, the patient’s serum was positive by both FT and PH and (b) the other two sera had positive confirmatory IgM tests and convalescent sera which converted from IgG negative to positive. In addition, all of the three latter discrepant samples had clinical history highly consistent with WNV. After resolution of the discrepancies, sensitivity of the FT test was 100% (5/5) and specificity 99% (89/90). 3.4. Clinical impact Generally physicians were contacted by telephone within 0–2 days of the test to communicate the results and to obtain their assessment of the clinical impact of the information. The vast majority of the time, physicians were contacted directly on the same day the test was performed. Alternatively, nurses were contacted if the physicians were not available. Fifty-five (58%) had clinical impact, 21 (22%) had undetermined impact, 12 (13%) were patient requests, 4 (4%) had no impact; 1 was done because the serum came from a potential donor of blood/tissue and on 2, the physician said the test was unnecessary but the test was ordered under the physician’s name (Table 1). Physicians generally were most appreciative of having the results in a more timely manner. Clinical impact for patients who tested positive included such reasons as (1) a patient was a candidate for a treatment study for WNV and (2) anti-viral therapy for Herpes simplex could be discontinued. Clinical impact for patients who tested negative was mainly that the information helped rule out WNV and set the workup in a different direction. In one case, the
124
J. Barenfanger et al. / Journal of Clinical Virology 34 (2005) 122–124
Table 1 Clinical impact of timely reporting of WNV serology Significance of test
Number of specimens (%)
Positive clinical impact Patient request Undetermined No impact Unnecessary Donor Total
55 (58) 12 (13) 21 (22) 4 (4) 2 (2) 1 (1) 95 (100)
patient was a candidate for an anti-dementia drug since WNV had been excluded.
4. Discussion Although there are many exceptions, most public health labs (by virtue of their construct as a tertiary reference laboratory) generally will not have the responsiveness or turnaround time of an onsite clinical laboratory. (The exceptions include such as expedited testing of rabies, hantavirus and SARS. In fact, in California WNV testing is often completed in a few days by a network of state and local public health laboratories and results are immediately relayed to clinicians.) However, it is possible that public health laboratories are used for tests intended to help manage patients (which is not the usual function of such public health labs). This study supports the concept that a rapid turn-around time for this test is clinically useful. Currently, two FDA-approved IgM Capture ELISA diagnostic kits are available to clinical laboratories for the serological diagnosis of West Nile Virus—one from Focus Technologies used in this study and another from PanBio Limited (Windsor, Australia). Our findings of the performance of the test are consistent with other studies. Moore et al. (2003) found a sensitivity of 99.6% and Prince et al. (2004) found a sensitivity and specificity of 100%. All of the FT negative samples had 100% concordance with PH test results. Only one of the four discrepant FT positive samples was a FT false positive (the patient had heterophile antibody which may cause a false positive for WNV). Heterophile antibodies present in human sera may react with proteins, typically immunoglobulin, from non-human species such as goats, horses, rabbits and mice. Although estimates of the prevalence of heterophile antibodies in normal sera vary from 0.5% to as high 20%, most think the prevalence is <2%. In rare circumstances, heterophile antibodies can be a cause of false positive results in serology tests. This is due to the fact that many laboratory tests utilize goat, rabbit or mouse sera as components of the reagents. If the absorbance test had been performed after the test was first determined to be positive (as the manufacturer recommends),
then the heterophile antibody would have been detected and the specimen would not have been determined to be positive. The other three discrepant FT positive samples were PH false negatives. Although Illinois was the hardest hit for WNV in 2002, it had substantially fewer cases at the time of this study, with only 54 documented cases and 1 death compared to 884 cases and 66 deaths the previous year (www.idph.il.us.gov). However, physicians were still most appreciative of notification of the test result, even when the vast majority of the results were negative. Just over half physicians responded in a manner suggesting that there was clinical impact for the information. This implies that it would be clinically useful to offer this test in a timely manner and that clinical diagnostic laboratories may not want to rely on testing from the public health laboratory in Illinois. Since rapid turn-around time is not a priority for most public health laboratories, this issue mostly likely pertains in other states as well. In one instance we described, the positive result of WNV enabled a physician to discontinue acyclovir. It is quite possible that patients with unknown encephalitidies are on anti-bacterial, anti-TB and anti-fungal medications as well. Therefore, a positive test result may impact management beyond just the use of acyclovir. Additionally, finding an etiologic agent is often very important to family members. In conclusion, the sensitivity of the FT test was 100% and specificity 99%, and this test was determined to have a clinical impact in 58% of the cases. As a result, this diagnostic laboratory offered WNV testing for three times a week during the peak mosquito season in 2004.
Acknowledgement This study was supported in part by Focus Technologies, Cypress, CA.
References Hogrefe WR, Moore R, Lape-Nixon M, Wagner M, Prince HE. The performance of West Nile Virus recombinant (preM/E)-based IgG and IgM ELISA for the detection of West Nile Virus and other flavivirus antibodies. J Clin Microbiol 2004;42:4641–8. Makar RS, Stowell CP. West Nile Virus: emerging infection in transfusion medicine. Clin Microbiol Newsl 2004;26:65–70. Moore RJ, Prince HE, Lape-Nixon M, Hogrefe WR. Comparison of the Focus Technologies West Nile Virus IgG and IgM kit results to public health serology results. In: Proceedings of the 52nd Annual Meeting of the American Society of Tropical Medicine and Hygiene; 2003. Prince H, Lape-Nixon MR, Moore RJ, Hogrefe WR. Utility of the focus technologies West Nile Virus IgM capture enzyme-linked immunosorbent assay for testing cerebrospinal fluid. J Clin Microbiol 2004;42:12–5 May 11, 2004, www.idph.il.us.gov.