Parvovirus B19 infection in medical students during a hospital outbreak

Transcription

1 Journal of Medical Microbiology (2004), 53, DOI /jmm Parvovirus B19 infection in medical students during a hospital outbreak DanielE.Noyola, 1 M.LourdesPadilla-Ruiz, 1 M.GuadalupeObregón-Ramos, 1 Patricia Zayas 2 and Beatriz Pérez-Romano 2 Correspondence Daniel E. Noyola cherpitel@hotmail.com 1 Microbiology Department, Facultad de Medicina, Universidad Autónoma de San Luis Potosí, San Luis Potosí, Mexico 2 Department of Immunology, Laboratorios Clínicos de Puebla, Puebla, Mexico Received 8 August 2003 Accepted 14 November 2003 From March to May 2002, a parvovirus B19 (B19) outbreak was identified at a general hospital that serves as a teaching facility for the Universidad Autónoma de San Luis Potosí, Mexico. Medical students attending the hospital presented with symptoms suggestive of B19 infection. Previous studies have suggested that apparent hospital-related B19 outbreaks may be a reflection of B19 infection in the community. A study was undertaken to assess whether exposure to the hospital was a risk factor for B19 infection and to determine to what extent medical students were infected during this outbreak. The incidence of B19 infection in medical students attending the teaching hospital during the outbreak (n ¼ 211) was determined and compared to students not attending the hospital (n ¼ 96). To assess if a community-wide outbreak had occurred, 80 blood donors were also evaluated for the presence of B19 antibodies. Acute B19 infection was identified in 40 of 119 (33. 6 %) susceptible students attending the hospital and in 20 of 47 (42. 6 %) susceptible students not attending the hospital. The frequency of acute infection among susceptible blood donors was lower (9.5 %) than in students, but higher than the rate expected during non-epidemic periods. Most infections (68. 3 %) were asymptomatic. Symptoms reported by infected subjects were not specific for B19 infection. Only 11.7 % of subjects with acute infection fulfilled the clinical surveillance definition used to detect cases during the outbreak. In conclusion, hospital exposure was not associated to increased risk of B19 infection among medical students. Medical students may be at increased risk for acquiring and transmitting B19 infection during outbreaks. INTRODUCTION Parvovirus B19 (B19) is distributed worldwide and is the cause of erythema infectiosum, usually in children (Anderson et al., 1983; Plummer et al., 1985; Nunoue et al., 1985; Chorba et al., 1986; Brown, 2000). Adults infected with B19 may present with fever, rash, arthralgias and arthritis. Infection with B19 also accounts for aplastic crises in patients with haemolytic anaemia, chronic anaemia in immunosuppressed patients, as well as some cases of chronic arthritis (Rao et al., 1983; Kelleher et al., 1984; White et al., 1985; Reid et al., 1985; Brown, 2000). If infection occurs during pregnancy, B19 may cause fetal loss, or hydrops fetalis (Brown, 2000). Infection with B19 is common during childhood and school personnel are at risk of acquiring this infection (Gillespie et al., 1990). Hospital workers have not been found to be at high risk of infection during endemic periods (Adler et al., 1993). However, hospital acquisition and transmission have been reported after exposure to patients Abbreviation: B19, parvovirus B19. with acute infection (Evans et al., 1984; Bell et al., 1989; Pillay et al., 1992; Farr et al., 1996). Most reported outbreaks have been limited to a specific ward of a hospital (Bell et al., 1989; Pillay et al., 1992; Farr et al., 1996). Dowell et al. (1995) reported that a B19 outbreak, which was identified in a maternity ward, was in fact a generalized outbreak throughout the community with similar infection rates within and outside the hospital. From 8 March to 14 May 2002, a B19 outbreak was identified at a general hospital (Hospital Central Dr Ignacio Morones Prieto, San Luis Potosí, Mexico). Several medical students attending the hospital reported symptoms suggestive of B19 infection. We are unaware of previous reports describing the risk of B19 infection among medical students on clinical services in the hospital environment. The objective of this study was to determine whether hospital exposure during the outbreak was associated with acute B19 infection in medical students. METHODS This was an observational, analytical study. The study was approved by & 2004 SGM Printed in Great Britain IP:

2 D. E. Noyola and others the Ethics Committee at the School of Medicine, Universidad Autónoma de San Luis Potosí, and written informed consent was obtained from participating subjects. Subjects. Medical students enrolled at the School of Medicine, Universidad Autónoma de San Luis Potosí, San Luis Potosí, Mexico. Third, fourth and fifth year medical students who attended the hospital during March, April and May 2002 were considered as exposed to B19 at the hospital. Third, fourth and fifth year medical students who were attending other hospitals, as well as first year medical students (who do not attend any hospital), were enrolled as control, non-exposed subjects. All medical students attending class during the second week of May 2002 were asked to complete a questionnaire that included demographical information, the presence of symptoms compatible with B19 infection during the previous months, as well as exposure to the hospital environment, and invited to participate in a serological survey. Those who accepted provided a single blood sample for serological testing. The blood samples were centrifuged, the serum was separated and stored at 20 8C until testing was performed. In addition, determination of anti-b19 antibodies was performed on sera from community blood donors who were sampled in May and June The sera from these subjects were obtained from a community blood bank unrelated to the hospital. Laboratory studies and definitions. Sera were tested for the presence of B19 antibodies. An ELISA was used to detect IgM (Parvovirus B19 IgM ì-capture ELISA; IBL-Hamburg) and IgG antibodies to B19 (Parvovirus B19 IgG ELISA; IBL-Hamburg) (Tolfvenstam et al., 1996). Testing was performed according to the manufacturer s recommendations. Subjects who tested positive for IgG antibodies with negative IgM antibodies were considered to have been infected prior to the outbreak and classified as immune to B19. Subjects who tested positive for IgM antibodies were considered to have an acute infection. Subjects without IgM- and IgG-specific antibodies were considered seronegative and not infected during the outbreak. Sample-size calculation. It was estimated, based on the literature, that 50 % of students would be immune to B19 (IgG-positive) and not susceptible to infection during the outbreak (Dowell et al., 1995). The annual seroconversion rate in health-care workers during non-epidemic periods has been reported to be 0. 5 % per year in susceptible persons (Adler et al., 1993). Thus, in the absence of a community outbreak, the number of susceptible subjects with acute infection (IgM-positive) at a given time would be at most 0. 5 %. Therefore, it would be expected that at most 0.25 % of the non-exposed subjects would be IgM-positive at a given time if the B19 outbreak was limited to the hospital. Seroconversion rates of at least 25 % have been reported in susceptible subjects during hospital epidemics (Bell et al., 1989; Pillay et al., 1992). Thus, it would be expected that 12.5 % of third, fourth and fifth year medical students exposed to the hospital environment would be IgM-positive during the outbreak. We expected to include exposed and control subjects on a 3 : 1 ratio due to the number of students that made up each group. Based on these estimates, it was determined that 224 subjects (56 non-exposed and 168 exposed) would be required to determine if exposure to the hospital environment was associated with B19 infection with 95 % confidence and 80 % power. Statistical analyses. Statistical analyses were performed using SPSS software. Continuous variables were compared using Student s t-test or Mann Whitney U-test according to data distribution. Categorical data were analysed using Fisher s exact test and the chi-square test with the continuity correction. All comparisons were two-sided and a P value less than or equal to 0.05 was considered as significant. Multivariate analyses using logistical regression were performed including factors that might be related to B19 infection, such as exposure to the hospital environment and exposure to children (less than 15 years old) at home. Variables that were not equally distributed between study groups were also included in the multivariate analyses as covariates. RESULTS AND DISCUSSION There were 477 students who completed the written questionnaire. Three-hundred-and-seven (64. 3 %) students participated in the serological survey and are included in this report. Demographical features of students participating and not participating in the serological survey are presented in Table 1. Variables associated with higher participation rates included being female, older age, being enrolled in the third and fifth year of medical school and attendance at the hospital. There were no other differences between subjects participating and not participating in the study. Demographical features, risk factors for B19 infection and symptoms consistent with B19 infection reported by students included in the serological survey exposed to the hospital and not exposed to the hospital are presented in Table 2. B19 immunity prior to the outbreak Overall, 141 (45. 9 %) students were considered to be immune prior to the outbreak (Table 3). Third year medical students showed the lowest rate of prior immunity (38. 5%) compared to all other students (48. 5 %), but this difference was not statistically significant (P ¼ 0. 16). There was no association between age or other demographical factors and prior immunity to B19, with the exception of the presence of children at home. Contrary to what we expected, students reporting the presence of children younger than 15 years of age at home were less likely to have prior immunity to B19 (40.2 %) compared to those without children at home (54.2 %; P ¼ 0. 02). B19 infection among susceptible students Among students susceptible to infection, 60 (36. 1 %) suffered an acute infection. The association between demographical features and exposure history with acute infection are presented in Table 3. Female students and students exposed to the operating room or the surgical wards were significantly more likely to develop acute infection on univariate analysis. Logistical regression analysis was performed including variables found to be significant on univariate analysis, exposure to the hospital and variables that were not distributed evenly between subjects exposed and not exposed to the hospital environment (age, exposure to B19 case, children at home, number of children at home). After multivariate analysis, the only factors that were significantly associated with acute B19 infection were being female (P ¼ ; adjusted odds ratio 2. 95; 95 % confidence interval ) and attending the surgical wards of the hospital (P ¼ ; adjusted odds ratio 4. 79; 95 % confidence interval ). Previous reports have shown a higher infection rate among women (Morgan- Capner et al., 1987). It has been suggested that women are 142 IP: Journal of Medical Microbiology 53

3 Parvovirus B19 outbreak in medical students Table 1. Demographical features of medical students participating and not participating in the serological survey of B19 infection Numbers in parentheses are either the percentage or the range. Participated in study (n 307) Did not participate in study (n 170) P value Female 163/211 (53.1) 62/170 (36.5) Single 297/306 (97.1) 162/170 (95.3) 0.46 Age in years (median) 21 (18 29) 20 (18 38), B19 clinical definition 20/307 (6. 5) 4/167 (2. 4) Contact with B19 case 122/290 (42. 1) 55/162 (34) Children (younger than 15 years old) 174/305 (57) 101/169 (59. 8) at home Number of children at home 1 (0 6) 1 (0 7) 0.59 (median) Household contact with a rash 12/235 (5.1) 5/120 (4.2) 0.89 Year of medical school attended: First year 87/307 (28.3) 79/170 (46.5), Third year 78/307 (25.4) 20/170 (11.8) Fourth year 45/307 (14.7) 42/170 (24.7) Fifth year 97/307 (31.6) 29/170 (17.1) Exposure to the hospital 211/307 (68. 7) 72/170 (42. 4), Exposure to the operating room 18/305 (5. 9) 9/170 (5. 3) Table 2. Demographical features and risk factors for exposure to B19 in medical students during a hospital outbreak Numbers in parentheses are either the percentage or the range. Exposed to hospital (n 211) Not exposed to hospital (n 96) P value Female 107/211 (50.7) 56/96 (58.3) 0.26 Single 203/211 (96.2) 94/95 (98.9) 0.34 Age in years (median) 22 (19 29) 19 (18 25), B19 clinical definition 17/211 (8.1) 3/96 (3.1) 0.17 Contact with B19 case 111/199 (55. 8) 11/91 (12. 1), Children (younger than 15 years old) 108/209 (51. 7) 66/96 (68. 8) at home Number of children at home 1 (0 6) 1 (0 5) (median) Household contact with a rash 9/155 (5.8) 3/80 (3.8) 0.71 Hospital ward: Paediatrics (fifth year students) 48/211 (22.7) Obstetrics (fifth year students) 47/211 (22.3) Surgery (fourth year students) 40/211 (19) Internal Medicine (third year 76/211 (36) students) Exposure to the operating room 18/210 (8. 6) more likely to be exposed to children from whom they may acquire the infection. However, we did not find an independent association between exposure to children at home and acute infection. It should be pointed out that we were only provided with information regarding the number of children at home and not the extent of contact with them, which may be a more reliable marker for the risk of acquiring an infection. Women, especially mothers, are more likely to have closer contact with children and, thus, may be at higher risk of acquiring B19 infection. The higher frequency of acute IP:

4 D. E. Noyola and others Table 3. B19 antibody status among medical students during a hospital outbreak Immune to B19 Considered susceptible Acute infection (percentage of susceptible) P value* All subjects studied (36.1) Gender: Female (44.2) 0.02 Male (25.4) Marital status: Single (36. 3) 1 Married (40) Age in years (median) B19 clinical definition: Yes (53.8) 0.23 No (34.6) Contact with B19 case: Yes (32) 0.51 No (38.3) Children (younger than 15 years old) at home: Yes (37.5) 0.5 No (31. 7) Number of children at home (median) Household contact with rash: Yes (60) 0.17 No (35.5) Hospital ward: Paediatrics (fifth year students) (31.8) Obstetrics (fifth year students) (33.3) Surgery (fourth year students) (60.9) Internal medicine (third year students) (20.8) Exposure to the hospital: Yes (33. 6) No (42. 6) Exposure to the operating room: Yes (70) 0.04 No (34) *Analysis includes only subjects susceptible to B19. infection in women is of importance, since pregnant women are at risk of delivering a newborn with hydrops or may have other adverse events during pregnancy (Anand et al., 1987). Opposite to what we expected, exposure to the hospital wards during the outbreak months was not associated with acute infection (P ¼ 0. 89), nor was any other factor. It has previously been shown that what may appear as nosocomial spread of B19 infection may be in fact a manifestation of a community outbreak that had not been perceived (Dowell et al., 1995). Ray et al. (1997) reported that health-care workers exposed to patients with transient aplastic crisis were as likely to suffer acute infection as non-exposed health-care workers, indicating that community acquisition of B19 infection may frequently account for perceived nosocomial transmission. The acute infection rate among susceptible students that we found (36. 1 %) is similar to that reported in hospital outbreaks by others. Bell et al. (1989) reported a 44 % conversion rate among susceptible hospital workers in close contact with patients with sickle-cell disease and acute B19- associated aplastic crisis. In that report, only close contacts to the affected patients were evaluated. Thirty-three per cent of susceptible subjects studied during another B19 outbreak at a children s ward were found to have acute infection (Pillay et al., 1992). Dowell et al. (1995) reported % acute infection rates among susceptible subjects inside and outside of the hospital during a community-wide B19 outbreak. Medical students attending the surgical wards had significantly higher rates of acute infection (62. 5 %) among susceptible subjects compared to students attending other wards, or not attending the hospital. This high rate of acute infection among students compares to the 67 % infection 144 IP: Journal of Medical Microbiology 53

5 Parvovirus B19 outbreak in medical students rate reported in susceptible household contacts of B19 IgMpositive erythema infectiosum cases (Chorba et al., 1986). Of interest, the first subjects to be identified with B19-associated symptoms were anaesthesiology residents, and the subject considered to be the index case for the hospital outbreak was an intern rotating on the surgical service. Our data did not identify factors that accounted for the higher infection rate in students attending the surgical wards. Whether closer contact among these students or exposure to a patient with prolonged viral excretion, such as an immunocompromised patient, accounted for this high infection rate was not determined. Miyamoto et al. (2000) described a B19 outbreak among hospital workers without known exposure to patients with erythema infectiosum or aplastic crisis. Their findings suggest that transmission among hospital staff may be responsible for some B19 outbreaks among health-care workers. In this study, we assessed the presence of infection at a single point in time; therefore, the number of students infected during the outbreak may have been underestimated. Subjects infected early in the course of the epidemic could have lost IgM antibodies by the time serum was obtained and may have been erroneously classified as immune prior to the outbreak. In addition, subjects infected late in the course of the outbreak may have not developed IgM antibodies at the time the study was performed and could have been classified as not infected erroneously. Detection of B19 DNA by the polymerase chain reaction may identify the presence of B19 earlier than IgM antibodies during the course of an acute infection, and may persist longer than IgM antibodies (Brown, 2000). The determination of viral DNA in serum may have helped to identify some subjects with acute infection and negative IgM antibodies. Unfortunately, we did not have assays for the detection of B19 DNA available. Since students were sampled in a short period of time and these limitations apply to both the exposed and non-exposed students, it is unlikely that our conclusions may be affected by these methodological limitations. Infection control measures Once the outbreak was identified, several infection control measures were put into place at the hospital. Information regarding the outbreak was provided to all services, and infection control measures, in particular hand-washing, were reinforced. Students and hospital staff were encouraged to wear masks. All cases fulfilling a clinical case definition for acute B19 infection were identified and sera were obtained for confirmation of acute infection. The number of new cases was assessed weekly until no new cases were identified. Pregnant staff were evaluated to assess the presence of acute infection and provided with paid leave until the outbreak had ceased. B19 infection in the community Because of the large number of students detected with acute B19 infection, evidence for community spread was sought. Specific B19 antibodies were determined in sera of subjects who donated blood at a major blood bank unrelated to the hospital. Sera from 80 donors were tested and acute infection was found in four (9.5 %) of 42 susceptible subjects (Table 4). This rate was considerably lower than that of susceptible medical students. Since a larger proportion of blood donors was male compared to students, logistical regression analysis including gender and age was performed. After this analysis, susceptible students were still more likely to have an acute infection (P ¼ 0. 03; adjusted odds ratio 9. 6; 95 % confidence interval ) than blood donors. Although the infection rate in blood donors was lower than that in medical students, it was higher than the rate that would be expected in the absence of a community outbreak. Kuo et al. (2002) tested 400 blood donors for the presence of B19 IgM antibodies without a single positive result. Heegaard et al. (2002) reported absence of B19 IgM antibodies in the sera of 153 healthy bone-marrow donors. Thus, our data suggest that a B19 outbreak was present throughout the community during the months that the outbreak was identified at the hospital. Clinical manifestations in students with acute infection Infection was asymptomatic in 41 (68. 3 %) affected students. Rates of asymptomatic infections of 46 % have been reported during endemic periods (Adler et al., 1993) and 65 % during a community outbreak (Gillespie et al., 1990). Symptoms reported by students with acute infection were not specific for B19 infection and included fever (8. 3 %), rash (10 %), arthralgias (23. 3 %), arthritis (10 %), myalgias (18. 3 %), cervical adenopathy (11. 7 %), malaise (35 %), anorexia (16. 7 %) and headache (25 %). Symptoms were severe enough to require medical attention in 10 % of those with acute infection. Only seven students (11. 7 %) of those with acute infection fulfilled the surveillance definition for acute B19 infection used to identify possible cases during the outbreak. Dowell et al. (1995) also found that clinical symptoms correlated poorly with acute infection during the outbreak they investigated. In summary, % of susceptible students suffered B19 Table 4. B19 infection in medical students and blood donors Numbers in parentheses represent percentages. Medical students (n 307) Blood donors (n 80) P value Age (median) 21 24, Female 163 (53.1) 10 (12.5), Previous infection 141 (45.9) 38 (47.5) 0.9 Susceptible: Not infected 106 (63.9) 38 (90.5) Acute infection 60 (36.1) 4 (9.5) IP:

6 D. E. Noyola and others acute infection during a local outbreak and women were more likely to present evidence of acute infection. Students attending the surgical wards of the hospital were at high risk of acquiring B19 infection, while students attending other wards were not at increased risk of B19 acquisition compared to students not attending the hospital at all. The high rates of acute infection among asymptomatic subjects perceived at low risk for exposure, such as the students not attending the hospital and blood donors, highlight the need to study subjects apparently not exposed to the outbreak to determine the extent to which the virus is present in the community. Our observations suggest that when a B19 outbreak occurs in the community there may be a high infection rate in young adults who are in close contact, such as medical students. ACKNOWLEDGEMENTS We thank Dr Alejandro Ruiz-Argüelles and Dr Ernesto Sánchez- Hermosillo for their help throughout the duration of this study. REFERENCES Adler, S. P., Manganello, A. M., Koch, W. C., Hempfling, S. H. & Best, A. M. (1993). Risk of human parvovirus B19 infections among school and hospital employees during endemic periods. J Infect Dis 168, Anand, A., Gray, E. S., Brown, T., Clewley, J. P. & Cohen, B. J. (1987). Human parvovirus infection in pregnancy and hydrops fetalis. N Engl J Med 316, Anderson, M. J., Jones, S. E., Fisher-Hoch, S. P., Lewis, E., Hall, S. M., Bartlett, C. L., Cohen, B. J., Mortimer, P. P. & Pereira, M. S. (1983). Human parvovirus, the cause of erythema infectiosum (fifth disease)? Lancet i, Bell, L. M., Naides, S. J., Stoffman, P., Hodinka, R. L. & Plotkin, S. A. (1989). Human parvovirus B19 infection among hospital staff members after contact with infected patients. N Engl J Med 321, Brown, K. E. (2000). Parvovirus B19. In Mandell, Douglas and Bennett s Principles and Practice of Infectious Diseases, 5th edn, pp Edited by G. E. Mandell, J. E. Bennett & R. Dolin. Philadelphia: Churchill Livingstone. Chorba, T., Coccia, P., Holman, R. C. & 10 other authors (1986). The role of parvovirus B19 in aplastic crisis and erythema infectiosum (fifth disease). J Infect Dis 154, Dowell, S. F., Török, T. J., Thorp, J. A., Hedrick, J., Erdman, D. D., Zaki, S. R., Hinkle, C. J., Bayer, W. L. & Anderson, L. J. (1995). Parvovirus B19 infection in hospital workers: community or hospital acquisition? J Infect Dis 172, Evans, J. P. M., Rossiter, M. A., Kumaran, T. O., Marsh, G. W. & Mortimer, P. P. (1984). Human parvovirus aplasia: case due to cross infection in a ward. Br Med J (Clin Res Ed) 288, 681. Farr, R. W., Hutzel, D., D Aurora, R. & Rugino, T. (1996). Parvovirus B19 outbreak in a rehabilitation hospital. Arch Phys Med Rehabil 77, Gillespie, S. M., Cartter, M. L., Asch, S., Rokos, J. B., Gary, G. W., Tsou, C. J., Hall, D. B., Anderson, L. J. & Hurwitz, E. S. (1990). Occupational risk of human parvovirus B19 infection for school and day-care personnel during an outbreak of erythema infectiosum. JAMA (J Am Med Assoc) 263, Heegaard, E. D., Petersen, B. L., Heilmann, C. J. & Hornsleth, A. (2002). Prevalence of parvovirus B19 and parvovirus V9 DNA and antibodies in paired bone marrow and serum samples from healthy individuals. J Clin Microbiol 40, Kelleher, J. F., Jr, Luban, N. L. C., Cohen, B. J. & Mortimer, P. P. (1984). Human serum parvovirus as the cause of aplastic crisis in sickle cell disease. Am J Dis Child 138, Kuo, S. H., Lin, L. I., Chang, C. J., Liu, Y. R., Lin, K. S. & Cheng, A. L. (2002). Increased risk for parvovirus B19 infection in young adult cancer patients receiving multiple courses of chemotherapy. J Clin Microbiol 40, Miyamoto, K., Ogami, M., Takahashi, Y., Mori, T., Akimoto, S., Terashita, H. & Terashita, T. (2000). Outbreak of human parvovirus B19 in hospital workers. J Hosp Infect 45, Morgan-Capner, P., Wright, J., Longley, J. P. & Anderson, M. J. (1987). Sex ratio in outbreaks of parvovirus B19 infection. Lancet ii, 98. Nunoue, T., Okochi, K., Mortimer, P. P. & Cohen, B. J. (1985). Human parvovirus (B19) and erythema infectiosum. J Pediatr 107, Pillay, D., Patou, G., Hurt, S., Kibbler, C. C. & Griffiths, P. D. (1992). Parvovirus B19 outbreak in a children s ward. Lancet 339, Plummer, F. A., Hammond, G. W., Forward, K., Sekla, L., Thompson, L. M., Jones, S. E., Kidd, I. M. & Anderson, M. J. (1985). An erythema infectiosum-like illness caused by human parvovirus infection. N Engl J Med 313, Rao, K. R. P., Patel, A. R., Anderson, M. J., Hodgson, J., Jones, S. E. & Pattison, J. R. (1983). Infection with parvovirus-like virus and aplastic crisis in chronic haemolytic anemia. Ann Intern Med 98, Ray, S. M., Erdman, D. D., Berschling, J. D., Cooper, J. E., Török, T. J. & Blumberg, H. M. (1997). Nosocomial exposure to parvovirus B19: low risk of transmission to healthcare workers. Infect Control Hosp Epidemiol 18, Reid, D. M., Reid, T. M. S., Brown, T., Rennie, J. A. N. & Eastmond, C. J. (1985). Human parvovirus-associated arthritis: a clinical and laboratory description. Lancet i, Tolfvenstam, T., Rudén, U. & Broliden, K. (1996). Evaluation of serological assays for identification of parvovirus B19 immunoglobulin M. Clin Diagn Lab Immunol 3, White, D. G., Woolf, A. D., Mortimer, P. P., Cohen, B. J., Blake, D. R. & Bacon, P. A. (1985). Human parvovirus arthropathy. Lancet i, IP: Journal of Medical Microbiology 53