Geographic risk factors for viral hepatitis and cytomegalovirus infection among United States Armed Forces blood donors K.C. HYAMS,E.R. CROSS, M.A. BIANCO,D.R. SOYK-SARTY, C.M.ROPER,W.L DAHUT,AND J.A. HOLMBERG In an effort to determine whether residence in a foreign country increases the risk of hepatitis B and C and cytomegalovirus (CMV) infection in United States (US) Armed Forces blood donors, 5719 volunteer donors at four US Navy blood banks were evaluated. Most participants were repeat donors (68%) and were young (mean age, 25 years), male (88%), and white (SO%), black (lo%), or Hispanic (7%). Birth outside of the United States was reported by 6 percent of subjects, and 34 percent had lived in a foreign count for more than 3 months. Twenty (0.3%) subjects had hepatitis B surface antigen (HBgg), and 100 (1-7%) had antibody to hepatitis B core antigen (anti-HBc). Thirly-four (0.6%) were repeatably reactive in enzyme-linked immunosorbent assay for antibody to hepatitis C virus (anti-HCV); 1 1 (0.2%) had anti-HCV in immunoblot assay. Of the 3484 donors tested for anti-CMV, 1117 (32.1%) were positive. When demographic characteristics were controlled for, both anti-HBc and anti-CMV seropositivities were independently associated in male blood donors with residence in the Philippines. Geographic factors were not associated with HBsAg and anti-HCV positivity. These findings indicate that the prevalence of serologic markers for viral hepatitis is low in military blood donors, but that residence in the Western Pacific is a risk factor for hepatitis B and CMV infection. TRANSFUSION 1992;32:644-647. Abbrevlatlons: ALT = alanlne amlnotransferase; antl-HBc = antlbody to hepatltls B core antigen; CMV = cytornegalovlrus; EUSA = enzymelinked lmmunosorbent assay; HBsAg = hepatltls B surface antlgen; HCV = hepatltls C vlrus; HIV-1 = human Immunodeflclency vlrus type 1; HTLV-1/11 = human T-lymphotroplc vlrus type I andlor 11; RlBA = recomblnant lmmunoblot assay.
STUDIESOF UNITEDSTATES(US) military populations have demonstrated an increased risk of viral hepatitis among troops stationed in regions of high endemicity.'-' Military blood donors who have traveled to foreign countries may therefore be at increased risk of hepatitis infection. However, few studies of military blood donors have been p e r f ~ r m e d ,and ~ * ~none has evaluated the geographic risk factors for viral hepatitis transmission. Geographic variations in the risk of cytomegalovirus (CMV)infection, another transfusionassociated infectious disease, have also been reported6 but have not been evaluated in military blood donors. In this study, the relation between residence in foreign countries and serologic markers for hepatitis B, hepatitis
C, and CMV infection was evaluated among blood donors at US military blood banks. Materials and Methods The study was conducted during 1990 and 1991 at four leading regional US Navy blood banks, located in Okinawa, Japan; San Diego, California; Bethesda, Maryland; and Portsmouth, Virginia. The volunteer blood donor populations for these centers are primarily composed of active-duty Navy and Marine Corps personnel who are recruited from nearby military installations during blood drives. All donors at military blood banks are screened according to eligibility criteria of the US Food and Drug Administration and the American Association of Blood Banks. The study was performed in accordance with the ethical standards of the Committee on Human Experimentation of the US Naval Medical Research Institute (Bethesda, MD). On the days when the study was conducted in each of the centers, all potential allogeneic blood donors were asked to participate. After providing voluntary informed consent, each study subject completed a standardized questionnaire that elicited basic demographic information and data related to travel outside of the United States. Subjects were asked whether they had ever lived or been on duty in a foreign country for more than 3 months. The results of routine laboratory tests conducted by the blood banks were obtained for each study subject. These tests included enzyme-linked immunosorbent assay (ELISA, Abbott Laboratories, North Chicago, IL) for hepatitis B surface antigen (HBsAg) and total antibody to hepatitis B core antigen (anti-HBc); ELISA with Western blot confirmation for human immunodeficiency virus type 1 (HIV-1) and human T-lym-
From the Epidemiology Division, United States Naval Medical Research Institute, Bethesda, Maryland; the Blood Bank Division, San Diego Naval Hospital, San Diego, California; the Blood Bank Branch, Portsmouth Naval Hospital, Portsmouth, Virginia; and the Department of Internal Medicine and the Blood Bank Branch, National Naval Medical Center, Bethesda, Maryland. Supported by the Naval Medical Research and Development Command, Naval Medical Command, National Capital Region, Bethesda, Maryland, Work Unit No. 3M162770AR122. The opinions and assertions contained herein are the private ones of the authom and are not to be construed as official or reflecting the view of the US Department of the Navy or Department of Defense. Received for publication October 28. 1991; revision received February 13, 1992, and accepted February 15, 1992.
644
TRANSFUSION 1992-Val. 32, No. 7
645
GEOGRAPHIC RISK AMONG BLOOD DONORS
photropic virus type I and/or I1 (HTLV-MI); latex agglutination test for total antibody to CMV (anti-CMV, Becton Dickinson, Gxkeysville, MD); nontreponemal antibody assays (RPRNDRL) confirmed by the fluorescent treponemal antibody-absorbed test; and analysis for serum alanine aminotransferase (ALT) levels by standard methods. Anti-CMV testing was not routinely performed on all blood donations but was used selectively to screen blood components for immunosuppressed recipients and to test group 0, Rh-negative blood. Initial ELISAs for antibody to hepatitis C virus (anti-HCV) were performed at the individual blood banks with first-generation commercial ELISA kits (Abbott or Ortho Diagnostics, Raritan, NJ). An aliquot of serum was obtained from repeatably reactive samples and tested with a second-generation immunoblot assay that used four HCV recombinant antigens (RIBA HCV Test System, Chiron Corporation, Emeryville, CA).’.* We compared proportions by using the x2 test with Yates’ correction or Fisher’s exact test. Multiple logistic regression analysis was performed with a statistical package (SPSSPC, SPSS Inc., Chicago, IL). Logistic models were developed by the maximum-likelihood method. We developed two final models, using the presence or absence of anti-HBc and antiCMV as the dichotomous outcome variable. Adjusted odds ratios were reported with 95-percent confidence intervals.
Results A total of 5719 blood donors were entered into the study: 1134 from Okinawa, 2311 from San Diego, 1147 from Bethesda, and 1127 from Portsmouth. The study population comprised 93.6 percent active-duty personnel, 0.7 percent military retirees, and 5.7 percent civilians. Most (68%) study participants were repeat donors. The mean age of participants was 25 years (range, 17-74); 88 percent were men. The racial and ethnic composition of the subject group was 80 percent white, 10 percent black, 7 percent Hispanic, and 3 percent “other.” Three hundred twenty-one subjects (5.6%) had been born outside of the United States,
and 1932 (33.8%)had lived in a foreign country for more than 3 months. Three percent of subjects (173) had received a blood component transfusion. Among the 5719 study subjects, 20 (0.3%)had HBsAg, 100 (1.7%)had anti-HBc, and 34 (0.6%)were repeatably reactive in ELISA for anti-HCV. Sera from 33 anti-HCV ELISA-reactive samples were available for supplemental testing by RlBA 11 (0.2%)were RIBA positive and 4 wen indeterminate. Of the 3484 samples tested for anti-CMV, 1117 (32.1%)were positive. ALT levels above the upper limit of normal (60KUL)were found in 351 participants (6.1%). By Western blot, one subject was positive for HIV-1 and none was positive for HTLV-UII. Nine subjects (0.2%) had a positive serologic test for syphilis. A similar percentage of subjects were positive for HBsAg, anti-HBc, and anti-HCV at the four blood bank centers and among active-duty military, retired military, and civilian donors. Nor was there a significant difference in the prevalence of these markers in first-time and repeat donors. There was no clear association between HBsAg positivity and various demographic characteristics or a histoxy of travel outside the United States (Tables 1 and 2). None of the HBsAg-positive subjects had ever received a blood transfusion. The 100 anti-HBc-positive subjects were more likely to be male, older, and foreign-born and to belong to black and “other” racial or ethnic groups (Table 1). Donors who had lived in a foreign country, particularly in Southeast Asia and the Western Pacific, had a higher prevalence of anti-HBc (Table 2). Moreover, a higher percentage of subjects who had received a blood component transfusion were anti-HBc positive (4.0 vs. 1.6%, p = 0.04). Among the 11 anti-HCV RIBA-positive subjects, all were male; 5 were white and 4 Hispanic (Table 1). There was no apparent association between anti-HCV seropositivity and foreign birth or travel (Table 2). One subject had received a blood component transfusion. Three RIBA-positive subjects had antiHBc, but none was positive for HBsAg. More anti-HCV ELISApositive subjects (14.7%)and RIBA-positive subjects (36.4%) than anti-HCV-negative subjects (6.1%)had ALT levels above
Table 1. Seroprevalence of hepatitis markers and anti-cytomegalovinrs (CMV) among mllltary blood donors Percentage positive' HBsAat Anti-HBc* Anti-HCVB Anti-CMV Gender Male Female Age (years)
0.4 (19/5054) 0.2 (1/665)
1.9 (96/5054) 0.6 (4/665)
0.2 (1115054) 0 (01665)
30.9 (97413152) 43.1 (143/332)
17-1 9 20-29 30-39 >39
0.5 (7/1425) 0.3 (11/3172) 0.3 (2/757) 0 (0/362)
1.4 (20/1425) 1.4 (45/3172) 3.6 (27/757) 2.2 (8/362)
0 0.3 0.4 0
26.6 (263/988) 32.0 (612/1913) 40.5 (1511373) 43.8 (911208)
(0/1425) (813172) (3/757) (01362)
Racelethnlcity 28.3 (801/2828) 0.1 (514547) 1.3 (61/4547) 0.3 (12/4547) White 49.2 (120/244) 3.0 (16/537) 0.2 (1/537) 0.9 (5/537) Black 44.0 (118/268) 1.0 (4/401) 2.2 (9/401) 0 (0/401) Hispanlc 58.8 (671114) 0.5 (1/183) 7.7 (14/183) 1.6 (3/183) Other Birthplace 0.2 (10/5398) 30.9 (102013297) 1.5 (8215398) 0.3 (17/5398) United States 51.9 (971187) 0.3 (11321) 5.6 (18/321) 0.9 (3/321) Other Number positivelnumber tested. Denominator totals varied slightly because of nonresponse to study questions, and anti-CMV testlng was not routinely performed on all blood donatlons. t Hepatitis B surface antigen. Includes HBsAg-positive subjects 5 Anti-hepatitis C virus-positive by recombinant immunoblot assay.
+
646 Table 2.
‘IRANSFUSION
HYAMS ET AL.
Vol. 32. No. 7-199l
Seroprevalence in military blood donors of hepatitis markers and anti-cytomegalovirus (CMV) by exposure in regions and countries outside of the United States
Percentagi positive* HBsAgt
Anti-HBcS
No
0.4 (1513774)
1.4 (53/3774)
Yes
0.3 (5/1932)
2.4 (4711932)
0
2.8 (6/215) 1.4 (7/487) 1.7 (21121) 2.0 (2/99) 6.0 (4167) 9.3 (171182) 6.8 (71103) 2.0 (2511250)
Antl-HCVO
Anti-CMV
Ever lived > 3 months outside of United States
Specific regions lived in for >3 months Central or South AmeridCaribbean Northern Europe Spain1ltaly1Greece Middle East/Afrlca Southeast Asia Philippines Guam JapanlOkinawall
0.2 0 0 0 1.1 0 0.2
(0/215) (11487) (0/121) (0/99) (0/67) (21182) (0/103) (2/1250)
0.2 3774) 0.2 1932)
(81
29.2 (80612760)
(3/
43.1 (308/714)
0.5 (11215) 0 (0/487) 0 (0/121) 0 (0/99) 0 (0167) 0 (01182) 0 (01103) 0.2 (21 1250) 0 (01114) 0 (01128)
0.9 (11114) 5.3 (61114) Korea Other 0 (01128) 0 (01128) Number poskivelnumber tested. t Hepatltls B surface antigen. Includes HBsAg-positive subjects. 8 Anti-hepatitls C virus-positlve by recombinant lmmunobiot assay. II Evaluated as positive In donors currently ilving in Okinawa if stationed in Okinawa for >3 months.
49.6 37.0 38.5 38.7 50.0 56.2 45.6 42.8
(641129) (901243) (30/78) (24/62) (20/40) (59/105) (31/68) (801187)
30.0 (12/40) 31.9 (22/69)
*
Table 3. Final logistlc regression models with variables independently assoclated with seroposltivlty for antibody to hepatitis B core antigen (antl-HBc) and cytomegalovirus (antiSignificant variables
CMV Odds ratio (95% Ci)*
Model with anti-HBc as outcome variable Male gender 3.49 (1.27-9.57) Age (by year) 1.03 (1.01-1.06) Racelethnlcity (referent white) 2.24 (1 -28-3.92) Black Hispanic 1.95 (0.95-3.99) Other 3.48 (1.72-7.03) Residence In 3.67 (139-7.1 1) the Philippines Model with antl-CMV as outcome variable Female gender 1.60 (1.25-2.04) Age @Y year) 1.03 (1.02-1.04) Racelethnicity (referent white) Black 2.41 (1.84-3.15) Hispanic 2.19 (1.68-2.85) Other 2.65 (1.74-4.04) Birth outside the 1.59 (1.14-2.22) United States Resldence in 1.77 (1.15-2.74) the Philippines * Confidence Interval.
p values 0.02 0.006 0.005 0.07
STUDIESOF UNITEDSTATES(US) military populations have demonstrated an increased risk of viral hepatitis among troops stationed in regions of high endemicity.'-' Military blood donors who have traveled to foreign countries may therefore be at increased risk of hepatitis infection. However, few studies of military blood donors have been p e r f ~ r m e d ,and ~ * ~none has evaluated the geographic risk factors for viral hepatitis transmission. Geographic variations in the risk of cytomegalovirus (CMV)infection, another transfusionassociated infectious disease, have also been reported6 but have not been evaluated in military blood donors. In this study, the relation between residence in foreign countries and serologic markers for hepatitis B, hepatitis
C, and CMV infection was evaluated among blood donors at US military blood banks. Materials and Methods The study was conducted during 1990 and 1991 at four leading regional US Navy blood banks, located in Okinawa, Japan; San Diego, California; Bethesda, Maryland; and Portsmouth, Virginia. The volunteer blood donor populations for these centers are primarily composed of active-duty Navy and Marine Corps personnel who are recruited from nearby military installations during blood drives. All donors at military blood banks are screened according to eligibility criteria of the US Food and Drug Administration and the American Association of Blood Banks. The study was performed in accordance with the ethical standards of the Committee on Human Experimentation of the US Naval Medical Research Institute (Bethesda, MD). On the days when the study was conducted in each of the centers, all potential allogeneic blood donors were asked to participate. After providing voluntary informed consent, each study subject completed a standardized questionnaire that elicited basic demographic information and data related to travel outside of the United States. Subjects were asked whether they had ever lived or been on duty in a foreign country for more than 3 months. The results of routine laboratory tests conducted by the blood banks were obtained for each study subject. These tests included enzyme-linked immunosorbent assay (ELISA, Abbott Laboratories, North Chicago, IL) for hepatitis B surface antigen (HBsAg) and total antibody to hepatitis B core antigen (anti-HBc); ELISA with Western blot confirmation for human immunodeficiency virus type 1 (HIV-1) and human T-lym-
From the Epidemiology Division, United States Naval Medical Research Institute, Bethesda, Maryland; the Blood Bank Division, San Diego Naval Hospital, San Diego, California; the Blood Bank Branch, Portsmouth Naval Hospital, Portsmouth, Virginia; and the Department of Internal Medicine and the Blood Bank Branch, National Naval Medical Center, Bethesda, Maryland. Supported by the Naval Medical Research and Development Command, Naval Medical Command, National Capital Region, Bethesda, Maryland, Work Unit No. 3M162770AR122. The opinions and assertions contained herein are the private ones of the authom and are not to be construed as official or reflecting the view of the US Department of the Navy or Department of Defense. Received for publication October 28. 1991; revision received February 13, 1992, and accepted February 15, 1992.
644
TRANSFUSION 1992-Val. 32, No. 7
645
GEOGRAPHIC RISK AMONG BLOOD DONORS
photropic virus type I and/or I1 (HTLV-MI); latex agglutination test for total antibody to CMV (anti-CMV, Becton Dickinson, Gxkeysville, MD); nontreponemal antibody assays (RPRNDRL) confirmed by the fluorescent treponemal antibody-absorbed test; and analysis for serum alanine aminotransferase (ALT) levels by standard methods. Anti-CMV testing was not routinely performed on all blood donations but was used selectively to screen blood components for immunosuppressed recipients and to test group 0, Rh-negative blood. Initial ELISAs for antibody to hepatitis C virus (anti-HCV) were performed at the individual blood banks with first-generation commercial ELISA kits (Abbott or Ortho Diagnostics, Raritan, NJ). An aliquot of serum was obtained from repeatably reactive samples and tested with a second-generation immunoblot assay that used four HCV recombinant antigens (RIBA HCV Test System, Chiron Corporation, Emeryville, CA).’.* We compared proportions by using the x2 test with Yates’ correction or Fisher’s exact test. Multiple logistic regression analysis was performed with a statistical package (SPSSPC, SPSS Inc., Chicago, IL). Logistic models were developed by the maximum-likelihood method. We developed two final models, using the presence or absence of anti-HBc and antiCMV as the dichotomous outcome variable. Adjusted odds ratios were reported with 95-percent confidence intervals.
Results A total of 5719 blood donors were entered into the study: 1134 from Okinawa, 2311 from San Diego, 1147 from Bethesda, and 1127 from Portsmouth. The study population comprised 93.6 percent active-duty personnel, 0.7 percent military retirees, and 5.7 percent civilians. Most (68%) study participants were repeat donors. The mean age of participants was 25 years (range, 17-74); 88 percent were men. The racial and ethnic composition of the subject group was 80 percent white, 10 percent black, 7 percent Hispanic, and 3 percent “other.” Three hundred twenty-one subjects (5.6%) had been born outside of the United States,
and 1932 (33.8%)had lived in a foreign country for more than 3 months. Three percent of subjects (173) had received a blood component transfusion. Among the 5719 study subjects, 20 (0.3%)had HBsAg, 100 (1.7%)had anti-HBc, and 34 (0.6%)were repeatably reactive in ELISA for anti-HCV. Sera from 33 anti-HCV ELISA-reactive samples were available for supplemental testing by RlBA 11 (0.2%)were RIBA positive and 4 wen indeterminate. Of the 3484 samples tested for anti-CMV, 1117 (32.1%)were positive. ALT levels above the upper limit of normal (60KUL)were found in 351 participants (6.1%). By Western blot, one subject was positive for HIV-1 and none was positive for HTLV-UII. Nine subjects (0.2%) had a positive serologic test for syphilis. A similar percentage of subjects were positive for HBsAg, anti-HBc, and anti-HCV at the four blood bank centers and among active-duty military, retired military, and civilian donors. Nor was there a significant difference in the prevalence of these markers in first-time and repeat donors. There was no clear association between HBsAg positivity and various demographic characteristics or a histoxy of travel outside the United States (Tables 1 and 2). None of the HBsAg-positive subjects had ever received a blood transfusion. The 100 anti-HBc-positive subjects were more likely to be male, older, and foreign-born and to belong to black and “other” racial or ethnic groups (Table 1). Donors who had lived in a foreign country, particularly in Southeast Asia and the Western Pacific, had a higher prevalence of anti-HBc (Table 2). Moreover, a higher percentage of subjects who had received a blood component transfusion were anti-HBc positive (4.0 vs. 1.6%, p = 0.04). Among the 11 anti-HCV RIBA-positive subjects, all were male; 5 were white and 4 Hispanic (Table 1). There was no apparent association between anti-HCV seropositivity and foreign birth or travel (Table 2). One subject had received a blood component transfusion. Three RIBA-positive subjects had antiHBc, but none was positive for HBsAg. More anti-HCV ELISApositive subjects (14.7%)and RIBA-positive subjects (36.4%) than anti-HCV-negative subjects (6.1%)had ALT levels above
Table 1. Seroprevalence of hepatitis markers and anti-cytomegalovinrs (CMV) among mllltary blood donors Percentage positive' HBsAat Anti-HBc* Anti-HCVB Anti-CMV Gender Male Female Age (years)
0.4 (19/5054) 0.2 (1/665)
1.9 (96/5054) 0.6 (4/665)
0.2 (1115054) 0 (01665)
30.9 (97413152) 43.1 (143/332)
17-1 9 20-29 30-39 >39
0.5 (7/1425) 0.3 (11/3172) 0.3 (2/757) 0 (0/362)
1.4 (20/1425) 1.4 (45/3172) 3.6 (27/757) 2.2 (8/362)
0 0.3 0.4 0
26.6 (263/988) 32.0 (612/1913) 40.5 (1511373) 43.8 (911208)
(0/1425) (813172) (3/757) (01362)
Racelethnlcity 28.3 (801/2828) 0.1 (514547) 1.3 (61/4547) 0.3 (12/4547) White 49.2 (120/244) 3.0 (16/537) 0.2 (1/537) 0.9 (5/537) Black 44.0 (118/268) 1.0 (4/401) 2.2 (9/401) 0 (0/401) Hispanlc 58.8 (671114) 0.5 (1/183) 7.7 (14/183) 1.6 (3/183) Other Birthplace 0.2 (10/5398) 30.9 (102013297) 1.5 (8215398) 0.3 (17/5398) United States 51.9 (971187) 0.3 (11321) 5.6 (18/321) 0.9 (3/321) Other Number positivelnumber tested. Denominator totals varied slightly because of nonresponse to study questions, and anti-CMV testlng was not routinely performed on all blood donatlons. t Hepatitis B surface antigen. Includes HBsAg-positive subjects 5 Anti-hepatitis C virus-positive by recombinant immunoblot assay.
+
646 Table 2.
‘IRANSFUSION
HYAMS ET AL.
Vol. 32. No. 7-199l
Seroprevalence in military blood donors of hepatitis markers and anti-cytomegalovirus (CMV) by exposure in regions and countries outside of the United States
Percentagi positive* HBsAgt
Anti-HBcS
No
0.4 (1513774)
1.4 (53/3774)
Yes
0.3 (5/1932)
2.4 (4711932)
0
2.8 (6/215) 1.4 (7/487) 1.7 (21121) 2.0 (2/99) 6.0 (4167) 9.3 (171182) 6.8 (71103) 2.0 (2511250)
Antl-HCVO
Anti-CMV
Ever lived > 3 months outside of United States
Specific regions lived in for >3 months Central or South AmeridCaribbean Northern Europe Spain1ltaly1Greece Middle East/Afrlca Southeast Asia Philippines Guam JapanlOkinawall
0.2 0 0 0 1.1 0 0.2
(0/215) (11487) (0/121) (0/99) (0/67) (21182) (0/103) (2/1250)
0.2 3774) 0.2 1932)
(81
29.2 (80612760)
(3/
43.1 (308/714)
0.5 (11215) 0 (0/487) 0 (0/121) 0 (0/99) 0 (0167) 0 (01182) 0 (01103) 0.2 (21 1250) 0 (01114) 0 (01128)
0.9 (11114) 5.3 (61114) Korea Other 0 (01128) 0 (01128) Number poskivelnumber tested. t Hepatltls B surface antigen. Includes HBsAg-positive subjects. 8 Anti-hepatitls C virus-positlve by recombinant lmmunobiot assay. II Evaluated as positive In donors currently ilving in Okinawa if stationed in Okinawa for >3 months.
49.6 37.0 38.5 38.7 50.0 56.2 45.6 42.8
(641129) (901243) (30/78) (24/62) (20/40) (59/105) (31/68) (801187)
30.0 (12/40) 31.9 (22/69)
*
Table 3. Final logistlc regression models with variables independently assoclated with seroposltivlty for antibody to hepatitis B core antigen (antl-HBc) and cytomegalovirus (antiSignificant variables
CMV Odds ratio (95% Ci)*
Model with anti-HBc as outcome variable Male gender 3.49 (1.27-9.57) Age (by year) 1.03 (1.01-1.06) Racelethnlcity (referent white) 2.24 (1 -28-3.92) Black Hispanic 1.95 (0.95-3.99) Other 3.48 (1.72-7.03) Residence In 3.67 (139-7.1 1) the Philippines Model with antl-CMV as outcome variable Female gender 1.60 (1.25-2.04) Age @Y year) 1.03 (1.02-1.04) Racelethnicity (referent white) Black 2.41 (1.84-3.15) Hispanic 2.19 (1.68-2.85) Other 2.65 (1.74-4.04) Birth outside the 1.59 (1.14-2.22) United States Resldence in 1.77 (1.15-2.74) the Philippines * Confidence Interval.
p values 0.02 0.006 0.005 0.07
