BRIEF REPORT needlesticks; nosocomial infection
Under-Reporting of Contaminated Needlestick Injuries in Emergency Health Care Workers Study hypothesis: There is considerable under-reporting of contaminated occupational needlestick and other sharp object injuries among emergency health care workers. Population: A convenience sample of emergency physicians, emergency nurses, and emergency medical technicians (EMTs). Methods: A survey instrument eliciting demographic and work-related factors was developed and administered; survey items included age, sex, occupation, years in occupation, n u m b e r of procedures performed per week, number of contaminated needlestick (and other "sharps") injuries recalled, and number of these injuries formally reported during the previous five years. Nonsegmented visual analog scales were used to assess eight attitudes possibly associated with nonreporting. Analysis was by analysis of variance and multiple linear regression with stepwise variable selection. Results: Two hundred fifty-nine subjects recalled 643 contaminated exposures during the five-year study period, but only 228 (35%) were formally reported. One or more injuries occurred in 55% of EMTs compared with 72% of nurses and 80% of physicians (P < .05). Physicians recalled a mean of 3.8 contaminated exposures, whereas nurses recalled 2.8 and EMTs recalled only 1.8 (P < .05). Physicians formally reported a mean of 0.26 exposures, whereas EMTs reported 0.85 and nurses reported 1.25 (P < .05). Physicians formally reported only one eighth of their injuries compared with EMTs and nurses, who each reported two thirds of these events (P < .05). Perception of risk, occupation, years in occupation, and concern about excessive paperwork were the most powerful predictors of low reporting rate (P < .05). Conclusion: Work-related contaminated sharp object injuries are underreported by emergency health care workers, especially emergency physicians. [Tandberg D, Stewart KK, Doezema D: Under-reporting of contaminated needlestick injuries in emergency health care workers. Ann Emerg Med January i991;20:66-70.]
Dan Tandberg, MD, FACEP Kenneth K Stewart, MD David Doezema, MD, FACEP Albuquerque, New Mexico From the Division of Emergency Medicine, University of New Mexico School of Medicine, Albuquerque. Received for publication April 23, 1990. Revision received July 13, 1990. Accepted for publication July 31, 1990. Supported in part by the Dean's Medical Student Research Fund, University of New Mexico School of Medicine. Address for reprints: Dan Tandberg, MD, FACER Division of Emergency Medicine, University of New Mexico School of Medicine, 620 Camino De Salud, NE, Albuquerque, New Mexico 87131.
INTRODUCTION Occupational injury by sharp objects such as needles, broken glass tubes, and scalpels is a risk for health care workers, mainly because of the possibility of contracting serious infectious diseases such as AIDS and hepatitis B.1 It has recently been suggested that those caring for emergency patients may be at particularly high risk, given the increasing prevalence of unsuspected human immunodeficiency virus {HIV-1) infection in the population and the propensity of such patients to seek care in emergency departments. 2 Because epidemiologic estimates of risk are necessarily based o~a analysis of only formally reported events, a high degree of under-reporting could lead to serious underestimation of the actual incidence of occupational transmission of HIV-1 and other pathogens. Efforts to reassure the health care profession that the risk of work-related HIV-1 transmission is small could result in further under=reporting of exposures and even greater underestimation of the actual risk. We conducted a survey of emergency physicians, emergency nurses, and emergency medical technicians {EMTs) to assess the reporting of these injuries and to determine factors and attitudes associated with failure to for-
20:1 January 1991
Annals of Emergency Medicine
66/89
N E E D L E S T I C K INJURIES Tandberg, Stewart & D o e z e m a
FIGURE 1. Comparison of total percutaneous contaminated sharp object exposures w i t h t h o s e f o r m a l l y reported. P h y s i c i a n s are m u c h l e s s likely than nurses and EMTs to report these events (P < .05).
Exposures
v. R e p o r t s Ex~x~es
250
Reacts 200
really report injuries. A preliminary survey instrument was designed and pretested on nurses and physicians not involved in the study. A variety of open-ended questions was used to elicit attitudes that might be associated with failure to report contaminated percutaneous sharp object exposures. Eight frequently recounted attitudes were identified as possible explanations for nonreporting and served as the basis of the study. The final version of the instrument elicited demographic and work-related factors, including age, sex, occupation, years in occupation, and number of procedures performed per week. It also contained items to ascertain the number of remembered c o n t a m i n a t e d needlestick injuries and the number of such injuries form a l l y reported to a p h y s i c i a n or employee health care facility during the previous five years. The questionnaire e x p l i c i t l y stated that nonneedlestick exposures to other contaminated sharp objects (such as scalpels or broken glass tubes) were to be included. Nonsegmented visual analog scales 100 m m in length were provided to allow expression of the extent of a g r e e m e n t or d i s a g r e e m e n t w i t h statements denoting the eight attitudes selected for study. 3-5 These attitudes included being too busy, too much paperwork, feeling the risk of infection was low, not wanting to appear careless, not wanting to know if there was a positive exposure, acceptance of exposure as part of one's job, being afraid of losing one's job, and not trusting the confidentiality of the health care system. Standard techniques were used to ensure clarity, simplicity, and n e u t r a l i t y of each questionnaire item and to encourage completeness of response.6y The questionnaire was administered to emergency physicians, emergency nurses, and EMTs in a variety of settings, including educational conferences, direct mail, and staff 90/67
150
d
METHODS
IO0
5O
0
EMTs
Nurses
1
Ph ysicia'~s
occ~PA~o.
TABLE 1. Means and 95% confidence intervals of demographic and workrelated variables for each occupational group No. in group Age (yr) Years in occupation Procedures (per week) Needlesticks (per five years) Reports (per five years) Reporting rate
EMTs 125 33 (32 35)
Nurses 88 38 (36-39)
Physicians 46 39 (37 40)
6.5 (5.7
9.8 (8.8-10.8)
8.0 (6.6
13 (10-17)
37 (33
17 (12-23)
1.8 (1.2 - 2.3)
2.8 (2.1 3.8)
3.8 (2.8 - 4.8)
0.85 (0.68 - 1.02) 0.67 (0.59 0.74)
1.25 (1,04 1.46) 0.66 (0,58- 0.73)
0.26 (0.02 - 0.56) 0.12 (0.01 0.22)
T.4)
41)
9.5)
Confidence inlervals are in parenlheses; those lhat de nol overlap are associaled with means signilicanlly different at the .05 level.
meetings at five public and private hospitals in the greater Albuquerque area. All responses were anonymous. The rate of reporting of contaminated needlestick injuries was calculated by dividing those formally reported by the number recalled by each subject. Attitude scores were measured to the nearest 0.5 ram. One-factor analysis of variance (ANOVA) was used to analyze differences in demographic and work-related items among the three occupational groups. 8 Statistical analysis of the association of reporting rate with all measured variables was by multiple linear regression with stepwise variable selection. 9 Multifactor ANOVA was used to evaluate attitude scale differences between groups of respondents. Correction for multiple comparisons and calculation of confidence intervals were made using the Tukey procedure.lO Two-tailed tests Annals of Emergency Medicine
and an c~ of .05 were used throughout. This study was reviewed and approved by the U n i v e r s i t y of N e w Mexico School of Medicine Human Research Review Committee. All responses were voluntary.
RESULTS Three hundred eighty-five questionnaires were distributed, and 9.84 (74%) were returned. Twenty-three were not analyzed because the respondents were not working primarily in emergency medicine, nursing, or prehospital care, and two were excluded because they were incomplete. Of the remaining 259 respondents, 125 were EMTs or paramedics, 88 were emergency nurses, and 46 were emergency physicians (Table 1). A total of 643 percutaneous exposures to sharp objects contaminated with patients' body fluids were re20:1 January 1991
NEEDLESTICK INJURIES Tandberg, S t e w a r t & D o e z e m a
TABLE 2. Predictors of low reporting rate Variable
F
P
"1 thought the risks werelow."
355
< 00001
Physician as occupation
33.1
< .00001
5.8
< .0001
Years in occupation "Reporting involves too much paperwork."
4.2
"1 was afraid of losing my job."
2.3
> .1
Age
1.0
> .5
'1 was much too busy at the time."
0.3
> .5
"1 do not trust the promise of confidentiality" "1 did not want to know the risk."
0.3 0.2
> .5 > .5
"Disagree"
0.2
> .5
0.1
> .5
'1 was afraid of appearing careless."
0.1
> .5
Nurse as occupation
0.1
> .5
Procedures per week 0.0 > .5 /: valuesand probal:ilitiesaredeterminedby linear regre~onwithstepwise variableselection,
called by subjects, but only 228 of these (35.5%) were formally reported. More than 55% of EMTs and paramedics had suffered at least one contaminated needlestick injury compared w i t h 72% of the nurses and 80% of the physicians. At least one exposure was recalled by 65% of all subjects. Significant differences between the three occupational groups were f o u n d for t h e m e a n n u m b e r of needlesticks, formally reported exposures, and rate of formal reporting (Table 1). Physicians recalled an average of 3.8 c o n t a m i n a t e d exposures during the previous five years, whereas EMTs recalled only 1.8 (P < .05). The m e a n for nurses was 2.8, which is not significantly different from the other two groups. Physicians formally reported an average of only 0.26 exposures during the previous five years, whereas EMTs reported 0.85 and nurses reported L25 (P < .05). Physicians reported only one eighth of their exposures compared w i t h EMTs and nurses, who each r e p o r t e d a p p r o x i m a t e l y t w o thirds of their exposures (Figure 1). Multivariate linear regression of all survey items with stepwise selection of variables revealed that perception of risk, occupation, years in occupation, and concern about the amount of paperwork were the most powerful
Stuck
v.
Stuck
90 f80
7o
Never
~r
J
Stuck
6O
< .0025
Sex "1 accept the risk as part of my job."
20:1 January 1991
Never
"Agree"
o
a:~-'eless low--risk occept busy poper~:~ afrold AT'n'B.E3ESCALE
predictors of low reporting rate (Table 2), Both forward and backward s e l e c t i o n s w e r e carried out w i t h identical results. Respondents who remembered at least one contaminated sharp object e x p o s u r e scored d i f f e r e n t l y t h a n those who had never been exposed on three of the eight attitude scales (Figure 2). Previously injured subjects were less concerned with appearing careless (P = .00003), less affected by being too busy (P = .004), and perceived the risk as being lower (P =
.011. Respondents who recalled formally reporting at least one contaminated needlestick injury scored differently than those who had never reported on three of the eight attitude scales (Figure 3). Reporting subjects were less concerned with being too busy (P = .0009), perceived the risk as being greater (P = .028), and expressed less denial regarding the risk involved (P = .037). Concern about appearing careless tended to be greater among those who had never reported an exposure but did not achieve statistical significance (P = .061). DISCUSSION Health care workers are at real risk for infection with at least 20 different pathogens from occupational injury with contaminated needles and other sharp objects. The Centers for Disease Control has estimated that there are 200 to 300 h e a l t h care worker deaths per year from occupationally acquired hepatitis B alone, n It is also clear that these employees are at low but genuine risk for HIV-1 Annals of Emergency Medicine
denld
privocy
FIGURE 2. Mean scores for the eight attitudes; comparison of s u b j e c t s never stuck with those stuck at least once by contaminated sharp objects. Stars denote means significantly different at the .05 level. infection from c o n t a m i n a t e d sharp objects.12-17 Epidemiologic studies have necessarily focused on formally reported occupational exposures. 18-21 However, it was noted even in the pre-AIDS era that under-reporting of these injuries may be very common despite intensive education and even mandatory requirements for reporting.22,23
We were interested in determining the incidence of under-reporting of contaminated "sharps" exposures in a high-risk work population and, if possible, to u n c o v e r those factors strongly associated with variations in reporting behavior. Personnel working in EDs and prehospital care settings are believed to be at especially high risk for this type of exposure 2 and for this reason might be expected to have higher-than-average reporting rates. This study demonstrated a high incidence of contaminated sharp object injury in the emergency care setting; more than 65% of our subjects recalled being percutaneously exposed to p a t i e n t s ' blood or b o d y fluids within the period studied. The finding that only 35% of these events were formally reported is surprising and suggests that for emergency caregivers current assessments of the risk of occupationally acquired infectious 66/91
NEEDLESTICK INJURIES Tandberg, Stewart & Doezema
FIGURE 3. Mean scores for the eight attitudes; comparison of subjects who never formally reported with those reporting at least once. Stars denote m e a n s significantly different at the .05 level.
Never "Disagree"
8o
"k
m
Reported
iiii
6O
92/69
v. R e p o r t e d
90
7O
diseases may be misleadingly low. T h e m o s t powerful predictor of low reporting rate was the subject's own perception of risk. Because epidemiologic e s t i m a t e s of risk analyze only f o r m a l l y reported injuries, ll,12,14,16-21 the high degree of under-reporting found in our study makes it very likely that there is serious underestimation of the real incidence of occupational transmission of HIV-1. Well-meant efforts to reassure the health professions that the risk is smallll,1a, lS,24,2s may actually lead to less reporting and further underappraisal of the true risk. This may also diminish compliance with " u n i v e r s a l b l o o d and b o d y fluid precautions." 11,25 Emergency physicians were m u c h less likely than nurses or EMTs to formally report their exposures, and these findings are consistent w i t h those of earlier, pre-AIDS era studies. ~6 One possible e x p l a n a t i o n is that m a n y physicians are self-employed and do not feel that there is m u c h personal gain in reporting, w h e r e a s n u r s e s and EMTs o f t e n work for large corporations and are perhaps more accustomed to "incident reporting" for self-protection. Another possible reason is that emergency physicians are usually not able to leave the ED during their shift; because most shifts take place outside of regular clinic hours, this further discourages formal reporting. Finally, it could be that emergency physicians actually believe the low published estimates of risk. The association between increased years in occupation and a low reporting rate has been described. 22,~3 Perhaps new workers are more responsive to employers' rules regarding formal reporting. Several of our subjects, h o w e v e r , d e s c r i b e d b e i n g sharply criticized during employee health visits for repeated needlestick exposures; if this practice is widespread, it could partially explain this finding. The perception that formal reporting involves an excessive amount of paperwork was strongly associated
Reported
5O 4O 3O 20 10
"Agree"
0
busy
Iow--rlskdenioicorelesspaper ofroid pdvCCYoccept AT'T]'I~OE SCALE
3
with a low reporting rate. It is noteworthy that subjects who had act u a l l y gone t h r o u g h the reporting process seemed less affected by this perception than those who had never reported. N e v e r t h e l e s s , efforts to streamline this process and make it less unpleasant would be expected to encourage reporting. Subjects w h o had suffered percutaneous exposure were less influenced by being too busy or appearing careless, suggesting that for some the experience is an emotionally charged one and results in some alteration of personal and professional priorities. Those who had actually gone through the formal reporting process were also less affected by being too busy and also exhibited less denial, perhaps reflecting the education and counseling that they received. Our study included emergency personnel from a wide range of busy work environments in a moderatesized city. It is possible that reporting rates in other occupational settings are different, but our study did not explore this. The only means of preventing occupational percutaneous transmission of serious communicabl.e diseases are i m m u n i z a t i o n , c h e m o p r o p h y l a x i s , a n d a v o i d a n c e of exposure; for HIV-1, only the last has proved effective. Despite efforts to reduce exposure and encourage reporting, the incidence of injury rem a i n s high, and o c c u r r e n c e s are strikingly under-reported. The attitudes found to be associated with low reporting rate suggest approaches for possible future intervention. Annals of Emergency Medicine
CONCLUSION Work-related contaminated n e e d l e s t i c k injuries are under-rep o r t e d by e m e r g e n c y h e a l t h care w o r k e r s in our c o m m u n i t y , especially e m e r g e n c y physicians. This suggests that current estimates of the incidence of occupationally acquired HIV-1 infection are low. Perception of risk, occupation, years in occupation, and concern about excessive paperwork were the most powerful predictors of low reporting rate. In the era of AIDS, it is worrisome that the surveillance of parenteral exposure in emergency health care personnel remains so imprecise. The authors appreciate the assistance from Sandra Mirabal in manuscript preparation; from Loren Cobb, PhD, who guided them in analysis of the data; and from Paul Gibson, who helped prepare the figures. REFERENCES 1. Kelen GD: H u m a n immunodeficiency virus and the emergency department: Risks and risk protection for health care providers. Ann Emerg Med 1990r19:242-248. 2. Kelen GD, Fritz S, Qaqish B, et al: Urtrecoginzed hum a n immunodeficiency virus infection in emergency department patients. N Eng] ] Med 1988~318:1645-1650. 3. Huskison EC: Visual analog scales, in Melzack R {ed): Pain Measurement and Assessment. N e w York, Raven Press, 1983, p 33-37. 4. Bond MR: Pain-Its Nature, Analysis and Treatment. N e w York, Churchill Livingstone, 1979, p 26-30. 5. Bond A, Lader M: The use of analogue scales in rating subjective feelings. Br J Med Psycho] 1974~47: 211-218. 6. Babbie ER: Survey Research Methods, ed 1. Belmont, California, Wadsworth Inc, 1973, p 131-156. 7. Hulley SB, C u m m i n g s SR: Designing Clinical Reseach, ed 1. Baltimore, Williams & wilkins, 1988, p 42-52. 8. Rice JA: Mathematical Statistics and Data Analysis,
20:1 January 1991
NEEDLESTICK INJURIES Tandberg, Stewart & Doezema
ed 1. Belmont, California, Wadsworth Inc, 1988, p 396-433.
nodeficiency syndrome among health care workers. N Engl J Med 1986;314:1127-1132.
9. Zar JH: Biostatistical Analysis, ed 2. Englewood Cliffs, New Jersey, Prentice-Hall, 1984, p 328-360.
15. Gerberding JL, Henderson DK: Design of rational infection control policies for human immunodeficiency virus infection. J Infect Dis 1987~156:861-864.
10. Kirk RE: Experimental Design, ed 2. Belmont, Califomia~ Wadsworth Inc, 1982, p 90-126. ~1. Centers for Disease Control: Guidelines for prevention of transmission of human immunodeficiency virus and hepatitis B virus to health-care and public safety workers. MMWR 1989~38:6S. I2. Becker CE, Cone JE, Geberding J: Occupational infection with h u m a n immunodefrciency virus (HIV). Arm Intern Med 1989;110:653-656. i3. Marcus R, CDC Cooperative Needlestick Surveillance Group: Surveillance of health care workers exposed to blood from patients infected with the human i m m u n o d e f i c i e n c y v i r u s . N EngI J Med 1988i 319:1118-1123. 14. McCray E: Occupational risk of the acquired immu-
20:1 January1991
16. Weiss SH, Saxinger WC, Rechtman D, et al: HTLVHI infection among health care workers: Association with needlestick injuries. JAMA 1985;254:2089-2094. 17. Hirsch MS, Wormser GP, Schooley RT, et al: Risk of nosocomial infection with human t-cell Iymphotropic vires lII (HTLV-III). N Engl J Med 1985;312:1-4. 18. H o c h r e i t e r MC, Barton LL: E p i d e m i o l o g y of needlestick injury in emergency medical service personnel. J Emerg Med 1988;6:9-12. 19. Ribner BS, Landry MN, Gholson GL, et al: Impact of a rigid, puncture resistant container system upon needlesrick injuries. Infect Control 1987;8:63-66. 20. Krasinski K, LaCouture i~, Holzman RS: Effect of changing needle disposal systems on needle puncture
Annals of Emergency Medicine
injuries. Infect Control 1987;8:59-62. 21. Jacobson JT, Burke JP, Conti MT: Injuries of hospital employees from needles and sharp objects. Infect Control 1983;4:100-102. 22. Hamory BH: Under-reporting of needlestick injuries in a university hospital. Am J Infect Control 1983;11: 174-177. 23. Hamory BH: Error: Percent in "under-reporting" of needlestiek injuries was under-reported. Am J lnfect Control 1984;i2:68. 24. Kuhls TL, Viker S, Parris NB, et ah Occupational risk of I-IIV, t-IBV and HSV-2 infection in health care personnel caring for AIDS patients. Am J Public Health 1987;77:1306-1309. 25. Baraff LJ, Talan DA: Compliance with universal precautions in a university hospital emergency department. Ann Emerg Med 1989;18:654-657. 26. McCormick RD, Maki DG: Epidemiology of needles t i c k i n j u r i e s in h o s p i t a l personnel. Am J Med 1981;70:928-932.
70/93
Under-Reporting of Contaminated Needlestick Injuries in Emergency Health Care Workers Study hypothesis: There is considerable under-reporting of contaminated occupational needlestick and other sharp object injuries among emergency health care workers. Population: A convenience sample of emergency physicians, emergency nurses, and emergency medical technicians (EMTs). Methods: A survey instrument eliciting demographic and work-related factors was developed and administered; survey items included age, sex, occupation, years in occupation, n u m b e r of procedures performed per week, number of contaminated needlestick (and other "sharps") injuries recalled, and number of these injuries formally reported during the previous five years. Nonsegmented visual analog scales were used to assess eight attitudes possibly associated with nonreporting. Analysis was by analysis of variance and multiple linear regression with stepwise variable selection. Results: Two hundred fifty-nine subjects recalled 643 contaminated exposures during the five-year study period, but only 228 (35%) were formally reported. One or more injuries occurred in 55% of EMTs compared with 72% of nurses and 80% of physicians (P < .05). Physicians recalled a mean of 3.8 contaminated exposures, whereas nurses recalled 2.8 and EMTs recalled only 1.8 (P < .05). Physicians formally reported a mean of 0.26 exposures, whereas EMTs reported 0.85 and nurses reported 1.25 (P < .05). Physicians formally reported only one eighth of their injuries compared with EMTs and nurses, who each reported two thirds of these events (P < .05). Perception of risk, occupation, years in occupation, and concern about excessive paperwork were the most powerful predictors of low reporting rate (P < .05). Conclusion: Work-related contaminated sharp object injuries are underreported by emergency health care workers, especially emergency physicians. [Tandberg D, Stewart KK, Doezema D: Under-reporting of contaminated needlestick injuries in emergency health care workers. Ann Emerg Med January i991;20:66-70.]
Dan Tandberg, MD, FACEP Kenneth K Stewart, MD David Doezema, MD, FACEP Albuquerque, New Mexico From the Division of Emergency Medicine, University of New Mexico School of Medicine, Albuquerque. Received for publication April 23, 1990. Revision received July 13, 1990. Accepted for publication July 31, 1990. Supported in part by the Dean's Medical Student Research Fund, University of New Mexico School of Medicine. Address for reprints: Dan Tandberg, MD, FACER Division of Emergency Medicine, University of New Mexico School of Medicine, 620 Camino De Salud, NE, Albuquerque, New Mexico 87131.
INTRODUCTION Occupational injury by sharp objects such as needles, broken glass tubes, and scalpels is a risk for health care workers, mainly because of the possibility of contracting serious infectious diseases such as AIDS and hepatitis B.1 It has recently been suggested that those caring for emergency patients may be at particularly high risk, given the increasing prevalence of unsuspected human immunodeficiency virus {HIV-1) infection in the population and the propensity of such patients to seek care in emergency departments. 2 Because epidemiologic estimates of risk are necessarily based o~a analysis of only formally reported events, a high degree of under-reporting could lead to serious underestimation of the actual incidence of occupational transmission of HIV-1 and other pathogens. Efforts to reassure the health care profession that the risk of work-related HIV-1 transmission is small could result in further under=reporting of exposures and even greater underestimation of the actual risk. We conducted a survey of emergency physicians, emergency nurses, and emergency medical technicians {EMTs) to assess the reporting of these injuries and to determine factors and attitudes associated with failure to for-
20:1 January 1991
Annals of Emergency Medicine
66/89
N E E D L E S T I C K INJURIES Tandberg, Stewart & D o e z e m a
FIGURE 1. Comparison of total percutaneous contaminated sharp object exposures w i t h t h o s e f o r m a l l y reported. P h y s i c i a n s are m u c h l e s s likely than nurses and EMTs to report these events (P < .05).
Exposures
v. R e p o r t s Ex~x~es
250
Reacts 200
really report injuries. A preliminary survey instrument was designed and pretested on nurses and physicians not involved in the study. A variety of open-ended questions was used to elicit attitudes that might be associated with failure to report contaminated percutaneous sharp object exposures. Eight frequently recounted attitudes were identified as possible explanations for nonreporting and served as the basis of the study. The final version of the instrument elicited demographic and work-related factors, including age, sex, occupation, years in occupation, and number of procedures performed per week. It also contained items to ascertain the number of remembered c o n t a m i n a t e d needlestick injuries and the number of such injuries form a l l y reported to a p h y s i c i a n or employee health care facility during the previous five years. The questionnaire e x p l i c i t l y stated that nonneedlestick exposures to other contaminated sharp objects (such as scalpels or broken glass tubes) were to be included. Nonsegmented visual analog scales 100 m m in length were provided to allow expression of the extent of a g r e e m e n t or d i s a g r e e m e n t w i t h statements denoting the eight attitudes selected for study. 3-5 These attitudes included being too busy, too much paperwork, feeling the risk of infection was low, not wanting to appear careless, not wanting to know if there was a positive exposure, acceptance of exposure as part of one's job, being afraid of losing one's job, and not trusting the confidentiality of the health care system. Standard techniques were used to ensure clarity, simplicity, and n e u t r a l i t y of each questionnaire item and to encourage completeness of response.6y The questionnaire was administered to emergency physicians, emergency nurses, and EMTs in a variety of settings, including educational conferences, direct mail, and staff 90/67
150
d
METHODS
IO0
5O
0
EMTs
Nurses
1
Ph ysicia'~s
occ~PA~o.
TABLE 1. Means and 95% confidence intervals of demographic and workrelated variables for each occupational group No. in group Age (yr) Years in occupation Procedures (per week) Needlesticks (per five years) Reports (per five years) Reporting rate
EMTs 125 33 (32 35)
Nurses 88 38 (36-39)
Physicians 46 39 (37 40)
6.5 (5.7
9.8 (8.8-10.8)
8.0 (6.6
13 (10-17)
37 (33
17 (12-23)
1.8 (1.2 - 2.3)
2.8 (2.1 3.8)
3.8 (2.8 - 4.8)
0.85 (0.68 - 1.02) 0.67 (0.59 0.74)
1.25 (1,04 1.46) 0.66 (0,58- 0.73)
0.26 (0.02 - 0.56) 0.12 (0.01 0.22)
T.4)
41)
9.5)
Confidence inlervals are in parenlheses; those lhat de nol overlap are associaled with means signilicanlly different at the .05 level.
meetings at five public and private hospitals in the greater Albuquerque area. All responses were anonymous. The rate of reporting of contaminated needlestick injuries was calculated by dividing those formally reported by the number recalled by each subject. Attitude scores were measured to the nearest 0.5 ram. One-factor analysis of variance (ANOVA) was used to analyze differences in demographic and work-related items among the three occupational groups. 8 Statistical analysis of the association of reporting rate with all measured variables was by multiple linear regression with stepwise variable selection. 9 Multifactor ANOVA was used to evaluate attitude scale differences between groups of respondents. Correction for multiple comparisons and calculation of confidence intervals were made using the Tukey procedure.lO Two-tailed tests Annals of Emergency Medicine
and an c~ of .05 were used throughout. This study was reviewed and approved by the U n i v e r s i t y of N e w Mexico School of Medicine Human Research Review Committee. All responses were voluntary.
RESULTS Three hundred eighty-five questionnaires were distributed, and 9.84 (74%) were returned. Twenty-three were not analyzed because the respondents were not working primarily in emergency medicine, nursing, or prehospital care, and two were excluded because they were incomplete. Of the remaining 259 respondents, 125 were EMTs or paramedics, 88 were emergency nurses, and 46 were emergency physicians (Table 1). A total of 643 percutaneous exposures to sharp objects contaminated with patients' body fluids were re20:1 January 1991
NEEDLESTICK INJURIES Tandberg, S t e w a r t & D o e z e m a
TABLE 2. Predictors of low reporting rate Variable
F
P
"1 thought the risks werelow."
355
< 00001
Physician as occupation
33.1
< .00001
5.8
< .0001
Years in occupation "Reporting involves too much paperwork."
4.2
"1 was afraid of losing my job."
2.3
> .1
Age
1.0
> .5
'1 was much too busy at the time."
0.3
> .5
"1 do not trust the promise of confidentiality" "1 did not want to know the risk."
0.3 0.2
> .5 > .5
"Disagree"
0.2
> .5
0.1
> .5
'1 was afraid of appearing careless."
0.1
> .5
Nurse as occupation
0.1
> .5
Procedures per week 0.0 > .5 /: valuesand probal:ilitiesaredeterminedby linear regre~onwithstepwise variableselection,
called by subjects, but only 228 of these (35.5%) were formally reported. More than 55% of EMTs and paramedics had suffered at least one contaminated needlestick injury compared w i t h 72% of the nurses and 80% of the physicians. At least one exposure was recalled by 65% of all subjects. Significant differences between the three occupational groups were f o u n d for t h e m e a n n u m b e r of needlesticks, formally reported exposures, and rate of formal reporting (Table 1). Physicians recalled an average of 3.8 c o n t a m i n a t e d exposures during the previous five years, whereas EMTs recalled only 1.8 (P < .05). The m e a n for nurses was 2.8, which is not significantly different from the other two groups. Physicians formally reported an average of only 0.26 exposures during the previous five years, whereas EMTs reported 0.85 and nurses reported L25 (P < .05). Physicians reported only one eighth of their exposures compared w i t h EMTs and nurses, who each r e p o r t e d a p p r o x i m a t e l y t w o thirds of their exposures (Figure 1). Multivariate linear regression of all survey items with stepwise selection of variables revealed that perception of risk, occupation, years in occupation, and concern about the amount of paperwork were the most powerful
Stuck
v.
Stuck
90 f80
7o
Never
~r
J
Stuck
6O
< .0025
Sex "1 accept the risk as part of my job."
20:1 January 1991
Never
"Agree"
o
a:~-'eless low--risk occept busy poper~:~ afrold AT'n'B.E3ESCALE
predictors of low reporting rate (Table 2), Both forward and backward s e l e c t i o n s w e r e carried out w i t h identical results. Respondents who remembered at least one contaminated sharp object e x p o s u r e scored d i f f e r e n t l y t h a n those who had never been exposed on three of the eight attitude scales (Figure 2). Previously injured subjects were less concerned with appearing careless (P = .00003), less affected by being too busy (P = .004), and perceived the risk as being lower (P =
.011. Respondents who recalled formally reporting at least one contaminated needlestick injury scored differently than those who had never reported on three of the eight attitude scales (Figure 3). Reporting subjects were less concerned with being too busy (P = .0009), perceived the risk as being greater (P = .028), and expressed less denial regarding the risk involved (P = .037). Concern about appearing careless tended to be greater among those who had never reported an exposure but did not achieve statistical significance (P = .061). DISCUSSION Health care workers are at real risk for infection with at least 20 different pathogens from occupational injury with contaminated needles and other sharp objects. The Centers for Disease Control has estimated that there are 200 to 300 h e a l t h care worker deaths per year from occupationally acquired hepatitis B alone, n It is also clear that these employees are at low but genuine risk for HIV-1 Annals of Emergency Medicine
denld
privocy
FIGURE 2. Mean scores for the eight attitudes; comparison of s u b j e c t s never stuck with those stuck at least once by contaminated sharp objects. Stars denote means significantly different at the .05 level. infection from c o n t a m i n a t e d sharp objects.12-17 Epidemiologic studies have necessarily focused on formally reported occupational exposures. 18-21 However, it was noted even in the pre-AIDS era that under-reporting of these injuries may be very common despite intensive education and even mandatory requirements for reporting.22,23
We were interested in determining the incidence of under-reporting of contaminated "sharps" exposures in a high-risk work population and, if possible, to u n c o v e r those factors strongly associated with variations in reporting behavior. Personnel working in EDs and prehospital care settings are believed to be at especially high risk for this type of exposure 2 and for this reason might be expected to have higher-than-average reporting rates. This study demonstrated a high incidence of contaminated sharp object injury in the emergency care setting; more than 65% of our subjects recalled being percutaneously exposed to p a t i e n t s ' blood or b o d y fluids within the period studied. The finding that only 35% of these events were formally reported is surprising and suggests that for emergency caregivers current assessments of the risk of occupationally acquired infectious 66/91
NEEDLESTICK INJURIES Tandberg, Stewart & Doezema
FIGURE 3. Mean scores for the eight attitudes; comparison of subjects who never formally reported with those reporting at least once. Stars denote m e a n s significantly different at the .05 level.
Never "Disagree"
8o
"k
m
Reported
iiii
6O
92/69
v. R e p o r t e d
90
7O
diseases may be misleadingly low. T h e m o s t powerful predictor of low reporting rate was the subject's own perception of risk. Because epidemiologic e s t i m a t e s of risk analyze only f o r m a l l y reported injuries, ll,12,14,16-21 the high degree of under-reporting found in our study makes it very likely that there is serious underestimation of the real incidence of occupational transmission of HIV-1. Well-meant efforts to reassure the health professions that the risk is smallll,1a, lS,24,2s may actually lead to less reporting and further underappraisal of the true risk. This may also diminish compliance with " u n i v e r s a l b l o o d and b o d y fluid precautions." 11,25 Emergency physicians were m u c h less likely than nurses or EMTs to formally report their exposures, and these findings are consistent w i t h those of earlier, pre-AIDS era studies. ~6 One possible e x p l a n a t i o n is that m a n y physicians are self-employed and do not feel that there is m u c h personal gain in reporting, w h e r e a s n u r s e s and EMTs o f t e n work for large corporations and are perhaps more accustomed to "incident reporting" for self-protection. Another possible reason is that emergency physicians are usually not able to leave the ED during their shift; because most shifts take place outside of regular clinic hours, this further discourages formal reporting. Finally, it could be that emergency physicians actually believe the low published estimates of risk. The association between increased years in occupation and a low reporting rate has been described. 22,~3 Perhaps new workers are more responsive to employers' rules regarding formal reporting. Several of our subjects, h o w e v e r , d e s c r i b e d b e i n g sharply criticized during employee health visits for repeated needlestick exposures; if this practice is widespread, it could partially explain this finding. The perception that formal reporting involves an excessive amount of paperwork was strongly associated
Reported
5O 4O 3O 20 10
"Agree"
0
busy
Iow--rlskdenioicorelesspaper ofroid pdvCCYoccept AT'T]'I~OE SCALE
3
with a low reporting rate. It is noteworthy that subjects who had act u a l l y gone t h r o u g h the reporting process seemed less affected by this perception than those who had never reported. N e v e r t h e l e s s , efforts to streamline this process and make it less unpleasant would be expected to encourage reporting. Subjects w h o had suffered percutaneous exposure were less influenced by being too busy or appearing careless, suggesting that for some the experience is an emotionally charged one and results in some alteration of personal and professional priorities. Those who had actually gone through the formal reporting process were also less affected by being too busy and also exhibited less denial, perhaps reflecting the education and counseling that they received. Our study included emergency personnel from a wide range of busy work environments in a moderatesized city. It is possible that reporting rates in other occupational settings are different, but our study did not explore this. The only means of preventing occupational percutaneous transmission of serious communicabl.e diseases are i m m u n i z a t i o n , c h e m o p r o p h y l a x i s , a n d a v o i d a n c e of exposure; for HIV-1, only the last has proved effective. Despite efforts to reduce exposure and encourage reporting, the incidence of injury rem a i n s high, and o c c u r r e n c e s are strikingly under-reported. The attitudes found to be associated with low reporting rate suggest approaches for possible future intervention. Annals of Emergency Medicine
CONCLUSION Work-related contaminated n e e d l e s t i c k injuries are under-rep o r t e d by e m e r g e n c y h e a l t h care w o r k e r s in our c o m m u n i t y , especially e m e r g e n c y physicians. This suggests that current estimates of the incidence of occupationally acquired HIV-1 infection are low. Perception of risk, occupation, years in occupation, and concern about excessive paperwork were the most powerful predictors of low reporting rate. In the era of AIDS, it is worrisome that the surveillance of parenteral exposure in emergency health care personnel remains so imprecise. The authors appreciate the assistance from Sandra Mirabal in manuscript preparation; from Loren Cobb, PhD, who guided them in analysis of the data; and from Paul Gibson, who helped prepare the figures. REFERENCES 1. Kelen GD: H u m a n immunodeficiency virus and the emergency department: Risks and risk protection for health care providers. Ann Emerg Med 1990r19:242-248. 2. Kelen GD, Fritz S, Qaqish B, et al: Urtrecoginzed hum a n immunodeficiency virus infection in emergency department patients. N Eng] ] Med 1988~318:1645-1650. 3. Huskison EC: Visual analog scales, in Melzack R {ed): Pain Measurement and Assessment. N e w York, Raven Press, 1983, p 33-37. 4. Bond MR: Pain-Its Nature, Analysis and Treatment. N e w York, Churchill Livingstone, 1979, p 26-30. 5. Bond A, Lader M: The use of analogue scales in rating subjective feelings. Br J Med Psycho] 1974~47: 211-218. 6. Babbie ER: Survey Research Methods, ed 1. Belmont, California, Wadsworth Inc, 1973, p 131-156. 7. Hulley SB, C u m m i n g s SR: Designing Clinical Reseach, ed 1. Baltimore, Williams & wilkins, 1988, p 42-52. 8. Rice JA: Mathematical Statistics and Data Analysis,
20:1 January 1991
NEEDLESTICK INJURIES Tandberg, Stewart & Doezema
ed 1. Belmont, California, Wadsworth Inc, 1988, p 396-433.
nodeficiency syndrome among health care workers. N Engl J Med 1986;314:1127-1132.
9. Zar JH: Biostatistical Analysis, ed 2. Englewood Cliffs, New Jersey, Prentice-Hall, 1984, p 328-360.
15. Gerberding JL, Henderson DK: Design of rational infection control policies for human immunodeficiency virus infection. J Infect Dis 1987~156:861-864.
10. Kirk RE: Experimental Design, ed 2. Belmont, Califomia~ Wadsworth Inc, 1982, p 90-126. ~1. Centers for Disease Control: Guidelines for prevention of transmission of human immunodeficiency virus and hepatitis B virus to health-care and public safety workers. MMWR 1989~38:6S. I2. Becker CE, Cone JE, Geberding J: Occupational infection with h u m a n immunodefrciency virus (HIV). Arm Intern Med 1989;110:653-656. i3. Marcus R, CDC Cooperative Needlestick Surveillance Group: Surveillance of health care workers exposed to blood from patients infected with the human i m m u n o d e f i c i e n c y v i r u s . N EngI J Med 1988i 319:1118-1123. 14. McCray E: Occupational risk of the acquired immu-
20:1 January1991
16. Weiss SH, Saxinger WC, Rechtman D, et al: HTLVHI infection among health care workers: Association with needlestick injuries. JAMA 1985;254:2089-2094. 17. Hirsch MS, Wormser GP, Schooley RT, et al: Risk of nosocomial infection with human t-cell Iymphotropic vires lII (HTLV-III). N Engl J Med 1985;312:1-4. 18. H o c h r e i t e r MC, Barton LL: E p i d e m i o l o g y of needlestick injury in emergency medical service personnel. J Emerg Med 1988;6:9-12. 19. Ribner BS, Landry MN, Gholson GL, et al: Impact of a rigid, puncture resistant container system upon needlesrick injuries. Infect Control 1987;8:63-66. 20. Krasinski K, LaCouture i~, Holzman RS: Effect of changing needle disposal systems on needle puncture
Annals of Emergency Medicine
injuries. Infect Control 1987;8:59-62. 21. Jacobson JT, Burke JP, Conti MT: Injuries of hospital employees from needles and sharp objects. Infect Control 1983;4:100-102. 22. Hamory BH: Under-reporting of needlestick injuries in a university hospital. Am J Infect Control 1983;11: 174-177. 23. Hamory BH: Error: Percent in "under-reporting" of needlestiek injuries was under-reported. Am J lnfect Control 1984;i2:68. 24. Kuhls TL, Viker S, Parris NB, et ah Occupational risk of I-IIV, t-IBV and HSV-2 infection in health care personnel caring for AIDS patients. Am J Public Health 1987;77:1306-1309. 25. Baraff LJ, Talan DA: Compliance with universal precautions in a university hospital emergency department. Ann Emerg Med 1989;18:654-657. 26. McCormick RD, Maki DG: Epidemiology of needles t i c k i n j u r i e s in h o s p i t a l personnel. Am J Med 1981;70:928-932.
70/93
