Needle-stick injuries in primary care in Wales
R. L. Atenstaedt, Specialist Registrar in Public Health Medicine, Honorary Lecturer in Public Health1,2,
S. Payne, Regional Director or Public Health (Retired)1
R. J. Roberts, Head of Vaccine Preventable Disease Programme1
I. T. Russell, Founding Professor of Public Health1,2
D. Russell, Senior Trial Statistician, NWORTH2
R. T. Edwards, Director of CEPhI1,3
1 National Public Health Service for Wales, UK
2 Institute for Medical and Social Care Research (IMSCaR), University of Wales, Bangor, UK
3 Centre for Economics and Policy in Health (IMSCaR), University of Wales, Bangor, UK
Address correspondence to Robert Atenstaedt, E-mail: robert.atenstaedt{at}nphs.wales.nhs.uk
Background Accidental needle-stick injuries (NSIs) are a hazard for health-care workers and for the general public.
Objectives To estimate the presentation rate of NSIs to general medical practices, their relation to practice characteristics, and review practice policies for managing NSIs.
Method Descriptive study using logistic regression analysis.
Results Annual rates of 2.73 (95% CI 2.08, 3.50) occupational NSIs per 100 clinical practice staff and 2.14 (95% CI 1.39, 3.13) non-occupational NSIs per 100 000 practice population were recorded. Stepwise logistic regressions showed that chance of a practice reporting at least one occupational NSI in previous five years was best predicted by being a single-handed practice (decreased odds). In contrast, the chance of a practice reporting at least one non-occupational NSI was best predicted by being a rural practice (increased odds). About one in five practices possessed no written policy on managing NSIs. Stepwise logistic regressions showed that the chance of a practice owning a NSI policy was best predicted by being located in an LHB area with a coastline (increased odds).
Conclusion NSIs are an important public health issue in Wales. We have tried to address the lack of guidance by developing new guidelines in Wales.
Keywords: general practice, needle-stick injury, primary care, public health
 |
Introduction |
|---|
 TOP Introduction Data and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
A needle-stick injury (NSI) is defined as ‘a penetrating wound with an instrument that is potentially contaminated with the body fluid of another person’.1 Accidental NSIs are an occupational hazard for health-care workers (HCWs) due to more than 100 000 injuries reported in UK hospitals annually.2 According to the World Health Organization, 16 000 Hepatitis C (HCV), 66 000 Hepatitis B (HBV) and 1000 cases of HIV may have occurred worldwide in the year 2000 among HCWs through their exposure to NSIs.3 Another major victim of NSIs is the general public,2 with discarded needles and syringes often found in parks, playing fields, beaches and alleyways.
A few studies have looked at the incidence of NSIs in the community4–10 and in practices1,11,12 and responses to the problem.13–16 However, none of these have estimated an incidence rate for NSIs in members of the public.
NSIs require appropriate evidence-based management to minimize the risk of infection. Guidelines on exposure to potential sources of HIV,17,18 HBV19 and HCV20 exist and have been adapted for use by some practices.
The objectives of this study were: to estimate the frequency of presentation of NSIs to general medical practices, both in practice staff (occupational) and members of the general public (non-occupational); to review practice policies for managing NSIs; and to investigate whether the incidence of NSIs and existence of policies are associated with practice characteristics. The underlying aim was to enhance the management of NSIs in primary care across Wales.
|
Data and methods |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
Primary care survey
The survey was completed in two stages: to all GP practices in North Wales in 200221 and to a random sample of half of the remaining practices in Wales in 2004 (after obtaining agreement to extend the project nationwide).
In the first phase, the sampling frame, a list of all 120 practices in North Wales, was compiled from the websites of the six Local Health Boards (LHBs) in the region. A 100% sample was selected, as the number of NSIs was believed to be small and as accurate a picture as possible was needed. There were no specific inclusion or exclusion criteria.
In the second phase of the study, a list of all practices in Wales was obtained from the Department of General Practice, Cardiff University. A 50% sample of practices in mid and South Wales was selected, stratified by LHB and number of principals (N = 190).
We compared the characteristics of responding and non-responding practices, including teaching status, single-handed or group, number of staff, list size, Townsend deprivation scores (obtained from the Department of Epidemiology and Public Health, Cardiff University) and location in rural or non-rural county (using information provided by the National Assembly for Wales).22
As information about NSIs presenting to GP practices was not available from routine data sources, it was decided to collect this information via a cross-sectional survey. There were no ‘off the shelf’ validated questionnaires that could be used for this survey. Therefore, a postal questionnaire for self-completion by practice managers was constructed using information gathered from the literature and from colleagues. The questionnaires sent were structured and used fixed, standardized questions, presented to respondents in the same way, with no variation in question wording.23 Each question offered several responses and asked respondents to tick the most appropriate. The questionnaire was in two parts. The first part was designed to elicit information on how many NSIs had presented to the practice in the previous five years. A data collection period of five years was chosen as NSIs are relatively infrequent and we judged this to be long enough to capture sufficient data. The second part sought to gather information on practice policies for managing NSIs. To test acceptability and validity, the questionnaire was piloted by practice managers out of area so as to avoid contamination of the study population. The piloting resulted in no changes to the format or wording of the instrument, as it was felt by respondents to have face validity in the sense of being clear and easy to complete and asking appropriate questions.
The definitive questionnaire was covered by a letter explaining the purpose of the study and recording that permission had been obtained from the relevant ethical committee. Also was included by a self-addressed Freepost envelope for return of completed questionnaires. Participants were initially given two weeks to reply. Those who had not returned their questionnaire within this period were sent another copy of the questionnaire and a reminder letter. If this was not returned within one week, practice managers were telephoned by the researcher. This call was to encourage further response, to ensure that the named people were still in post and to give practice managers the opportunity to complete the questionnaire ‘over the phone’.
Analysis
The completed questionnaires were coded, entered into Excel (Version 5) and analysed by importing the Excel file into SPSS (Version 10). Tables and charts were generated by Excel and SPSS to present the main results. Weighting was used to allow for the fact that all practices in North Wales were sampled, compared to only half in the rest of Wales. Practice clinical staff numbers and list populations were used as denominators for calculating NSI rates.
Because a very high proportion of practices had no NSIs (especially non-occupational), transformation, bootstrapping or non-parametric tests were not applicable. Hence, the main analysis used presence or absence of at least one NSI rather than rates and compared this with practice characteristics.
The statistical analyses were conducted in two steps: univariate analyses followed by a multivariate analysis using logistic regression. In the univariate analyses, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated with three dichotomous dependent variables (whether or not practice had at least one occupational NSI, had at least one non-occupational NSI and owned an NSI policy). Significance was assessed with chi-square tests. As practice characteristics are correlated, stepwise logistic regression was used to determine the combination of practice characteristics that best predicted the occurrence of NSIs or existence of a NSI policy. Practice variables used in the regression were Teaching, Research, Number of partners, Rural, Previous needle-stick injury, Coastal, Length of coast and Townsend deprivation score. Stepwise logistic regressions were done for the same three dependent variables as in the univariate analysis. Each logistic regression was then rerun with only the significant variables to minimize the loss due to missing responses.
|
Results |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
A total of 310 practices across Wales were surveyed; 249 (80%) responded.
Comparison of participants with non-participants
Although there was a suggestion that non-responding practices were more often single-handed, non-teaching and less likely to be involved in research, there was no significant difference between respondents and non-respondents in terms of the main characteristics except for Townsend (deprivation) score. Those who responded were significantly less likely to be in a deprived area than those who did not respond (median Townsend score for responders –0.479, non-responders 0.457, P = 0.036).
NSIs and ownership of NSI policies
Sixty-three percent of GP practices had at least one NSI present to them in the previous five years; 15% registered two, 5% three, 5% four and 9% five or more.
The number of NSIs recorded converts to annual rates of 
2.73 (95% CI 2.08, 3.50) occupational NSIs per 100 clinical practice staff and 2.14 (95% CI 1.39, 3.13) non-occupational NSIs per 100 000 practice population.
The majority [83% (207/249)] of practices had a written policy or guideline in place to deal with NSIs.
Table 1 shows the relationship between practice characteristics and recorded NSIs and the ownership of NSI policies. Table 2 shows significant practice variables in a logistic regression to predict whether a practice had at least one occupational NSI, had at least one non-occupational NSI and owned an NSI policy.
|
|
|
Discussion |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
Main findings
A response rate of 75% for postal surveys is generally accepted as good.23 The overall response rate to the survey was high at 80%, although this was not consistent across the 22 Local Health Boards, with a borderline significant relationship found between response and Townsend score (with those less deprived practices more likely to respond). This high response rate may be due to the tripartite approach used, which becomes increasingly personalized with each reminder. Initially, the questionnaire to each participant is un-named. The first reminder is more personalized with a name used and a handwritten signature. The second reminder utilizes a personal phone call to promote response. This method is efficient as it restricts more time-consuming data collection to those ‘hard to get to subjects’. The questionnaire design may have also improved the response rate. Pilot study respondents commented that it was short and did not take long to complete. In addition, NSI is a topic that seems to have captured the imagination of practice managers.
Rate of occupational NSIs and practice characteristics
Of the six practice variables, three were significantly associated with whether the practice had at least one occupational NSI: teaching status, research status and number of partners. When all variables were introduced into a stepwise logistic regression, two of the three variables remained significant, as can be seen in Table 2. Being a single-handed practice (decreased odds of having at least one NSI) had most effect but being a teaching practice (increased odds of having at least one occupational NSI) made a small but significant improvement to the prediction.
The Townsend deprivation score for individual practices was not significantly associated with the rate of occupational NSIs (r = –0.021, P = 0.750) and did not contribute to the logistic regression.
Rate of non-occupational NSIs and practice characteristics
Of the six practice variables, only one was significantly associated with whether the practice had at least one non-occupational NSI: rural status. The logistic regression showed that no other variable significantly improved the prediction of NSI, as can be seen in Table 2. Being a rural practice was associated with increased odds of having at least one non-occupational NSI.
The Townsend deprivation score for individual practices was not significantly associated with the rate of non-occupational (r = –0.039, P = 0.549) NSIs, and did not contribute to the logistic regression.
Guideline ownership and practice characteristics
Of the six practice variables, three were significantly associated with whether the practice owned a guideline on managing NSIs: number of partners, whether practice had sustained a previous NSI and was in an LHB area with a coastline. In the logistic regression, only two of the three variables remained significant, as can be seen in Table 2. Being a practice located in an LHB area with a coastline (increased odds of owning an NSI guideline) had most effect but being a single-handed practice (decreased odds of having a NSI guideline) made a small but significant improvement to the prediction.
Interpretation
For occupational NSIs, practices which recorded at least one NSI in the previous five years were significantly more likely to be group practices and teaching practices. Although the first finding is rather obvious given the greater number of staff in multi-partner practices, Table 1 shows that the number of NSIs per 100 clinical practice staff for group practices is approximately double that in single-handed practices (2.98 compared with 1.42). Group practices have a higher throughput of patients, undertake more interventions such as phlebotomy or might have better systems in place for recording NSIs. Teaching practices are more likely to have doctors in training.
The association between levels of deprivation and health inequalities has been long established. We might have expected that deprivation would act to increase both occupational and non-occupational NSIs. Social deprivation is linked to greater levels of ill health including intravenous drug abuse. In addition, illicit drug users are often drawn to areas of high deprivation. However, we did not find this to be the case. This negative result could reflect the use of a proxy marker for deprivation.
Our discovery that practices that had reported at least one NSI in the previous five years are more likely to be located in the rural counties of Wales is likely to be due to the large number of NSIs which result from farming practice.
The study found that the majority [83% (207/249)] of practices had a written guideline in place to deal with NSIs. However, almost one in five practices did not have a policy for managing NSIs and this is a public health concern. Practices owning NSI policies were significantly less likely to be single practices than group practices. This is to be expected, because of the greater manpower capacity of group practices. The fact that those with NSI policies are significantly more likely to be located in an LHB area with a coastline may be because beaches are often a location for discarded needles, and may prompt the ownership of a guideline. The variable ‘sustained previous NSI’ was significant in the univariate analysis, with practices reporting a previous NSI more likely to own a policy. However, the association became non-significant in the logistic regression, meaning that it is sufficiently explained by single-handed and costal.
What is already known on this topic
The rate of NSIs in practice staff recorded in a previous study in New Zealand (22 NSI per 100 staff in six months or 44 in one year) was much higher than that in this study.1 The lower rate in Wales might reflect tighter infection control practices or poorer reporting and/or recording systems. The ownership of NSI management policies recorded in this survey was higher than that in the previous studies described when figures of 38,12 51,1 6315 and 66%24 were reported. The high response rate in this survey is consistent with that in some other studies undertaken in general practice on this subject (have varied from 39,15 46,16 69,13 74.524 and 82%1).
What this study adds
Annual rates of 2.73 (95% CI 2.08, 3.50) occupational NSIs per 100 clinical practice staff and 2.14 (95% CI 1.39, 3.13) non-occupational NSIs per 100 000 practice population were recorded as presenting to general medical practices in Wales; this is the first study to give a rate for NSIs in the general public.
Limitations of this study
This descriptive study has succeeded in collecting robust quantitative data from GP practices in Wales on the incidence of NSIs and ownership of NSI policies; it is the first to calculate estimates for the general public. However, this study had a number of limitations. Only a small pilot study was conducted due to time constraints. Cross-sectional numerical estimates: precise data, perhaps obtained prospectively over a few years, would permit more exacting estimates of the incidence of NSIs. Data on NSIs were only obtained by practice and not by individual GP or practice nurse, which would have made it possible to explore other associations, for example, between the age of GPs and the number of NSIs sustained. It was also not possible to check the quality of the data received by practices, with poor recording systems leading to an under-estimation of NSIs. In many cases, the number of NSIs, especially in the general public, was estimated by GPs rather than derived from patient records or directly from the public themselves. The five-year time period would also make the data prone to recall bias. However, the small number of NSIs reported by individual practices might perhaps serve to minimize this form of bias. In addition, it is more likely that an event such as a NSI has been forgotten rather than inaccurately remembered as having occurred. This would cause the number of NSIs to be underestimated. If time had allowed, it would have been preferable to visit practices and validate the data personally. Another potential cause of under-estimation of NSIs is under-reporting of sharps injuries. The latter is well recognized, with a British study finding that only 9% of doctors and 46% of midwives questioned had reported occupationally acquired NSIs.25 Furthermore, some NSIs in the general public would present to A&E departments, which are not recorded in this study. It was not possible to estimate any cost or resource use of NSIs, which could be included in a future study.
A note of caution must be applied when interpreting the statistics in this study. In analysing the data, multiple comparisons were performed and so there is an increased chance of a variable yielding a positive result by chance. The logistic regression has partially addressed this shortcoming.
|
Conclusion |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
The study suggests that NSIs are an important public health issue in Wales. The overall rate of NSIs among clinical staff was substantial. The rate of NSI in practice staff was considerably higher than that in the general public. This is to be expected, as a non-occupational community NSI is an uncommon occurrence. We do not, unfortunately, have another general public rate of NSI from the literature to compare this with. About one in five Welsh practices surveyed possessed no written policy on the management of NSIs, which is of concern. We have tried to address this problem by developing new guidelines that will be discussed in another paper.
|
Funding |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
The authors thank the National Public Health Service for Wales for funding this study.
|
Acknowledgments |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
Also all the practices who diligently filled in questionnaires and returned them and the Departments of General Practice and Epidemiology and Public Health, Cardiff University, for providing information on GP practice characteristics. Contributors: RA formulated the original hypothesis, designed the methodology, carried out the initial analysis and wrote the first draft of the paper. SP, RR, IR and RTE helped to refine the hypothesis, design the methodology and suggested variables to study. DR devised and supervised the stepwise logistic regression analysis. All authors helped edit the final draft and approved its contents. RA is guarantor. Competing interests: None declared. Ethical approval: Local Research Ethics Committees in Wales confirmed that this study did not require formal Research Ethics Committee approval. This research earned Dr Atenstaedt the Young Epidemiologists Prize from the Royal Society of Medicine.
|
References |
|---|
|
TOPIntroductionData and methodsResultsDiscussionConclusionFundingAcknowledgmentsReferences |
|---|
- Lum D, Mason Z, Meyer-Rochow G, et al. Needle stick injuries in country general practice. NZ MJ (1997) 110(1041):122–5.
- Sharps injury information. [cited 11 March 2006]. Available at: < http://www.isips.org/sharps.shtml>.
- Pruss-Ustun A, Rapiti E, Hutin Y. Estimation of the global burden of disease attributable to contaminated sharps injuries among health-care workers. Am J Ind Med (2005) 48(6):482–90.[CrossRef][Web of Science][Medline]
- Philipp R. Community needlestick accident data and trends in environmental quality. Public Health (1993) 107(5):363–9.[CrossRef][Web of Science][Medline]
- Wyatt JP, Robertson CE, Scobie WG. Out of hospital needlestick injuries. Arch Dis Child (1994) 70(3):245–69.
[Abstract/Free Full Text] - Nourse CB, Charles CA, McKay M, et al. Childhood needlestick injuries in the Dublin metropolitan area. Irish Med J (1997) 90(2):66–9.
- Philipp R, Pond K, Rees G. Medical wastes found on coastline are increasing. BMJ (1994) 309:471.
[Free Full Text] - Philipp R, Pond K, Rees G. Research and the problems of litter and medical wastes on the UK coastline. B J Clin Practice (1994) 51(3):164–8.
- Russel FM, Nash MC. A prospective study of children with community-acquired needlestick injuries in Melbourne. J Paediatr Child H (2002) 38(3):322–3.[CrossRef][Web of Science][Medline]
- Pepper S. Sharps and needle-stick injuries in supported housing and emergency accommodation. J R Soc Health (2002) 122(2):74–5.[Web of Science][Medline]
- Bowden FJ. Needle-stick injuries in primary care. Aust Prescriber (2001) 24(4):98–100.
- Thurn J, Willenbring K, Crossley K. Needlestick injuries and needle disposal in Minnesota physicians' offices. Am J Med (1989) 86(5):575–9.[CrossRef][Web of Science][Medline]
- Caird A, Wann C. Are Irish General Practitioners minimising infection risk from sharps? Irish Med J (1995) 88(4):132–4.
- McCall L, Maher T, Piterman L. Preventive health behaviour among general practitioners in Victoria. Aust Fam Phys (1999) 28(8):854–7.[Medline]
- Miller KE, Krol RA, Losh DP. Universal precautions in the family physician's office. J Fam Practice (1992) 35(2):163–8.[Web of Science][Medline]
- Sneddon J, Ahmed S, Duncan E. Control of infection: a survey of general medical practices. J Public Health Med (1997) 19(3):313–9.
[Abstract/Free Full Text] - UK Health Departments. Guidelines on Post-Exposure Prophylaxis for Health Care Workers Occupationally Exposed to HIV (1997) Department of Health.
- Babl FE, Cooper ER, Kastner B, et al. Prophylaxis against possible human immunodeficiency virus exposure after nonoccupational needlestick injuries or sexual assaults in children and adolescents. Arch Pediatr Adolesc Med (2001) 155(6):680–2.
[Abstract/Free Full Text] - PHLS Hepatitis Subcommittee. Exposure to hepatitis B virus: guidance on post-exposure prophylaxis. CDR Review (1992) 2(9):98–101.
- Ramsay ME. Guidance on the investigation and management of occupational exposure to hepatitis C. CDPH (1999) 2(4):258–62.
- Atenstaedt R, Roberts R, Russell I, et al. Needlestick injuries presenting to GP practices across North Wales. CDPH (2004) 7(2):151.
- National Assembly for Wales. A Statistical focus on Rural Wales. (2001) [cited January 10 2006]. Available at: <http://www.wales.gov.uk/keypubstatisticsforwales/content/publication/compendia/2001/sb49-2001/sb49-2001.htm>.
- Bowling A. Research Methods in Health: Investigating Health and Health Services (1997) Buckingham: Open University Press.
- White RR, Smith JM. Infection control in general practice: results of a questionnaire survey. J Public Health Med (1995) 17(2):146–9.
[Abstract/Free Full Text] - Burke S, Madan I. Contamination incidents among doctors and midwives: reasons for non-reporting and knowledge of risks. Occup Med (1997) 47(6):357–60.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||