Blunt versus sharp suture needles for preventing percutaneous exposure incidents in surgical staff

Abstract

Background

Surgeons and their assistants are especially at risk of exposure to blood due to glove perforations and needle stick injuries during operations. The use of blunt needles can reduce this risk because they don't penetrate skin easily but still perform sufficiently in other tissues.

Objectives

To determine the effectiveness of blunt needles compared to sharp needles for preventing percutaneous exposure incidents among surgical staff.

Search methods

We searched MEDLINE and EMBASE (until May 2011), CENTRAL, NHSEED, Science Citation Index Expanded, CINAHL, Nioshtic, CISdoc, PsycINFO, and LILACS (until September 2010).

Selection criteria

Randomised controlled trials (RCTs) of blunt versus sharp suture needles for preventing needle stick injuries among surgical staff measured as glove perforations or self‐reported needle stick injuries.

Data collection and analysis

Two authors independently assessed study eligibility and risk of bias in trials and extracted data. We synthesized study results with a fixed‐effect model meta‐analysis.

Main results

We located 10 RCTs involving 2961 participating surgeons performing an operation in which the use of blunt needles was compared to the use of sharp needles. Four studies focused on abdominal closure, two on caesarean section, two on vaginal repair and two on hip replacement. On average, a surgeon that used sharp needles sustained one glove perforation in three operations. The use of blunt needles reduced the risk of glove perforations with a relative risk (RR) of 0.46 (95% confidence interval (CI) 0.38 to 0.54) compared to sharp needles. The use of blunt needles will thus prevent one glove perforation in every six operations.

In four studies, the use of blunt needles reduced the number of self‐reported needle stick injuries with a RR of 0.31 (95% CI 0.14 to 0.68). Because the force needed for the blunt needles is higher, their use was rated as more difficult but still acceptable in five out of six studies.

The quality of the evidence was rated as high.

Authors' conclusions

There is high quality evidence that the use of blunt needles appreciably reduces the risk of exposure to blood and bodily fluids for surgeons and their assistants over a range of operations. It is unlikely that future research will change this conclusion.

Plain language summary

The use of blunt needles compared to sharp needles for preventing needle stick injuries in surgical staff

Surgeons and their assistants are especially at risk of needle stick injuries during operations. This can lead to infection with HIV or other blood‐borne viruses. The use of blunt needles is proposed to prevent needle stick injuries. We reviewed the literature to evaluate the preventive effect of blunt needles compared to sharp needles on needle stick injuries among surgical staff. We searched multiple medical databases (to May 2011). We included studies if they were randomised controlled trials (RCTs) of blunt versus sharp suture needles for preventing needle stick injuries among surgical staff. We located 10 RCTs with 2961 operations in which blunt needles were compared to sharp needles. Six studies focused on abdominal operations, two on vaginal repair and two on hip replacement. On average, a surgeon that used sharp needles sustained one glove perforation per three operations. The use of blunt needles reduced the risk of glove perforations by 54% (95% confidence interval 46% to 62%) compared to sharp needles. The use of blunt needles in six operations will thus prevent one glove perforation. In four studies the use of blunt needles also reduced the number of self‐reported needle stick injuries by 69% (95% confidence interval 14% to 68%). Even though surgeons reported that the force needed for the blunt needles was higher, their use of the needles was still rated as acceptable in five out of six studies. We concluded that there is high quality evidence that the use of blunt needles appreciably reduces the risk of contracting infectious diseases for surgeons and their assistants over a range of operations by reducing the number of needle stick injuries. It is unlikely that future research will change this conclusion.

Summary of findings

Summary of findings for the main comparison. Blunt needles compared to sharp needles for preventing percutaneous exposure injuries.

Blunt needles compared to sharp needles for preventing percutaneous exposure injuries
Patient or population: surgical staff Intervention: blunt needles Comparison: sharp needles
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	sharp needles	blunt needles
glove perforations per surgeon per operation	Study population		RR 0.46 (0.38 to 0.54)	2961 (10 studies)	⊕⊕⊕⊕ high
	293 per 1000	135 per 1000 (111 to 158)
	Low risk population
	20 per 1000	9 per 1000 (8 to 11)
	High risk population
	750 per 1000	345 per 1000 (285 to 405)
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio;
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.

Background

Healthcare workers are at risk of acquiring infectious diseases through exposure at work. Exposure to blood or bodily fluids from infected patients can lead to infection with hepatitis B (HBV), hepatitis C (HCV) and HIV, among other pathogens. These are serious viral infections that may cause a chronic disease process or initiate cancer and eventually lead to death. According to the model of Pruss‐Ustun et al, 16,000 HCV, 66,000 HBV and 1000 HIV infections may have occurred worldwide among healthcare workers in the year 2000 due to their occupational exposure to blood and bodily fluids (Pruss‐Ustun 2005). The World Health Organization (WHO) reports that two million healthcare workers across the world experience percutaneous exposure to infectious diseases each year (WHO 2007). The operating room is a special environment in the healthcare setting with a high risk of sharps injuries (Watt 2008; Watt 2010).

Description of the condition

The risk of acquiring an infection is proportional to the prevalence of the infection in the patient population. Thus, in areas where hepatitis B is endemic and HIV is prevalent, such as in Africa, the risks are much higher than in Western Europe or North America. This situation has a significant impact on the health of the workers and also on the healthcare system as a whole. The transmission of occupational blood‐borne infectious diseases leads to absenteeism, morbidity and, in some cases, mortality among healthcare workers, which ultimately leads to a reduction in workforce and consequently affects patients' quality of care and safety. Because of the risk of acquiring an infectious disease at work, healthcare workers may also suffer from psychological stress, which affects both their work and personal life (Fisman 2002; Sohn 2006). There is also the financial burden imposed on hospitals due to occupational exposure to blood‐borne diseases, which includes costs related to blood tests, treatment, outpatient visits and lost working hours.

Description of the intervention

Exposure to blood or bodily fluids is also called percutaneous exposure and happens most often when healthcare workers are injured with sharp needles or instruments, or when blood or body fluids are splashed during medical interventions or accidents. These incidents are called sharps or needle stick injuries, or percutaneous exposure incidents (PEI). The actual causes of a PEI are multi‐factorial and include elements such as, but not limited to, types of devices and procedures, lack of access to or availability of personal protective equipment for the healthcare workers, sub‐optimal use of personal protective equipment, professional inexperience and lack of training and education on infection control and occupational health principles, improper management of sharps, poor organisational climate, high workload and fatigue, working alternative shifts, and high mental pressure and subjective perception of risk (Akduman 1999; Ansa 2002; Clarke 2002; Doebbeling 2003; Fisman 2007; Ilhan 2006; Oh 2005; Orji 2002; Roberts 1999; Smith 2006; Smith 2006b; Wallis 2007). Most of these causes can be addressed by specific interventions. In the operating theatre, one of the specific risks is a needle stick injury caused by suture needles. This accounts for more than half of all percutaneous injuries that occur in the operating room (Meyers 2008). Also, according to the American College of Surgeons, suture needles pose the greatest risk of sharp injuries to surgeons and scrub personnel (ACS 2011). The reporting of these injuries in surgical settings remained quite stable in the US between 1993 and 2003, in spite of the introduction of legislation requiring the use of safety‐engineered sharp devices in 2000. The lack of adoption of safer devices such as blunt suture needles is blamed for the lack of improvement (Jagger 2008; Jagger 2011).

How the intervention might work

There are several possibilities to prevent infection from PEI. For hepatitis B, vaccination has been successful (Chen 2005) but vaccination is not yet possible for HCV or HIV (Mast 2004). Therefore, exposure reduction remains the main preventive strategy. In general, there are several ways of reducing or eliminating exposure, such as elimination of hazards at the source (for example, elimination of unnecessary injections) or along the path (for example, safer medical devices or workplace practices, use of personal protective equipment) (Ellenbecker 1996; Roelofs 2003). For surgical staff, the prevention of sharp injuries related to sharp suture needles could be overcome by using blunt or taper point needles (ACS 2011; Monz 1991; NIOSH 2008). There is also evidence from laboratory studies that gloves give better protection against punctures with tapered needles than with sharp needles (Lefebvre 2008).

Why it is important to do this review

Blunt needles have been advocated for a long time (Wright 1993). It is important to know if this preventive intervention is effective. Some studies indicate that blunt needles would reduce PEIs by more than 50% without leading to less operative dexterity, whereas others argue that blunt needles simply don't work (Nordkam 2005; Wilson 2008). However, there is currently no review that gives a systematic summary of the available evidence on the effectiveness of blunt needles.

Objectives

To determine the effectiveness of blunt needles compared to sharp needles for preventing percutaneous exposure incidents among surgical staff.

Methods

Criteria for considering studies for this review

Types of studies

Randomised clinical trials (RCTs) and cluster‐randomised trials (c‐RCTs).

Types of participants

Persons working in the operation theatre that are exposed to the risk of percutaneous injuries with suture needles.

Types of interventions

Blunted suture needles compared to sharp needles. We defined blunt needles as suture needles that have a rounded blunt point and that are circular in diameter and that can be either curved or straight. Sharp needles are suture needles that have a tapered point and that can be either circular in diameter or square with cutting edges and that can be either curved or straight.

Types of outcome measures

Primary outcomes

Exposure of healthcare workers to contaminated blood or bodily fluids was our primary outcome measure. Exposure can be observed either as self‐reported needle stick injury or glove perforations. Many needle stick injuries and glove perforations are not noticed by the surgeon. A glove perforation means that the skin is exposed to blood or bodily fluids. A needle stick injury means that the skin has been perforated and the wound has been exposed to blood or bodily fluids. A needle stick injury is therefore a more serious exposure than a glove perforation. The occurrence is, however, proportional. In other words, an increase in glove perforations will be accompanied by an increase in percutaneous needle stick injuries. Therefore, we took both occurrences as a measure of outcome. Because it is easy for glove perforations to go unnoticed in situ, the perforations must be observed by means of a water or air test. In the water test the gloves are filled with water and the release of a jet‐stream of water is taken as a sign that there is a hole in the glove. In the air test, the gloves are filled with air and then submersed in water and air bubbling up to the surface indicates a hole in the glove. The combination of both tests was shown to detect all glove perforations that were made on purpose with 19, 21 and 23 gauge needles (Smith 1988; Smith 1990). The water test has been reported to detect only 80% of holes made on purpose (Brough 1988). For needle stick injuries, there are no objective tests so we took all self‐reports of needle stick injuries as an outcome.

Secondary outcomes

We included satisfaction with, or ease of use of, the needles.

Search methods for identification of studies

The search was part of a larger search for all interventions to prevent PEI in healthcare personnel and was intended to also include non‐randomised studies as we also searched for interventions that are more difficult to randomise such as safe work practices.

Electronic searches

First, search terms for PEI were applied. These terms for PEI were then combined with the recommended search strings for randomised trials and for non‐randomised studies. For randomised clinical trials and controlled clinical studies, we used the search strategy that has been developed by Robinson and Dickersin (Robinson 2002). For finding non‐randomised studies we used the sensitive search strategy for occupational health intervention studies (Verbeek 2005).

The strategy was used to search MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library) and the Occupational Safety and Health Review Group Specialised Register, NHSEED, Science Citation Index Expanded, CINAHL, Nioshtic, CISdoc, LILACS and PsycINFO (up until September 2010).

The search strategies for the databases are described in Appendix 1.

We did an additional search for the blunt needle intervention in MEDLINE and EMBASE (up until 27 April 2011), which we detail in Appendix 2.

Searching other resources

We searched the databases of the WHO, UK National Health Service (NHS) and www.med.virginia.edu/epinet. In addition, we checked the references of the included articles.

Data collection and analysis

Selection of studies

Using the inclusion and exclusion criteria, two authors working independently screened the identified titles and abstracts of the references that resulted from the search strategy for potential studies. We obtained the full‐text articles of the references that appeared to meet the inclusion criteria. Disagreements between authors were resolved by discussion. A third author (JV) was consulted if disagreements persisted.

Data extraction and management

Two authors independently extracted the data into a form. The form included essential study characteristics about the participants, the interventions, primary and secondary outcomes and results. We also intended to note any adverse events and the sponsorship of the study, but none were reported. The risk of bias of the studies was also independently assessed by two authors. In case of disagreements, a consensus method was used to come to a conclusion. A third author was consulted if disagreement persisted. We did not mask trial names because we feel that this would not increase validity.

Assessment of risk of bias in included studies

For the assessment of risk of bias in studies, we applied the risk of bias tool as provided in RevMan. We used the items on randomisation, allocation concealment, blinding of participants and outcome assessors, incomplete outcome data and selective outcome reporting, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Studies that had more than four out of the seven items rated as low risk of bias were rated as having low risk of bias, following the advice of the Cochrane Back Review Group (Furlan 2009).

Measures of treatment effect

Authors reported the outcome of their studies in many different ways. We assumed that the most valid estimate of the risk of exposure for a member of the operating staff was provided by the number of holes in gloves used by one surgeon or one assistant during one operation. It would have been more precise if 'operation' could have been defined as 'the number of hours engaged in an operation of average difficulty', comparable to a number of person‐years at risk, but the data were not sufficient to calculate this. We intended to treat the results of all trials as being dichotomous, even though needle stick injuries can be sustained more than once, because the resulting risk ratios are easier to interpret. However, the perforation rates in some studies were so high that the use of risk ratios was not feasible anymore. Therefore we calculated the natural logarithm (ln) of the rate ratios and their standard errors from the number of glove perforations and the number of surgeon‐operations in an Excel sheet, as recommended by the Cochrane Handbook (Higgins 2011). The ln rate ratios and the standard errors were used as input in RevMan where we combined them using the generic inverse variance method.

Unit of analysis issues

We intended to calculate the design effect for studies that employed a cluster‐randomised design but that did not make an allowance for the design effect. In six studies the unit of randomisation was the operation but the outcome was measured for one surgeon only. Therefore we assumed that there was no unit of analysis issue in these cases. In four studies there was more than one person who could sustain needle stick injuries in one operation. Therefore, these studies were effectively cluster‐randomised trials even though the authors did not design them as such (Mingoli 1996; Nordkam 2005; Sullivan 2009; Wilson 2008). There were, however, not enough details to calculate the effect of the clustering of the surgeons per operation. We calculated perforations in gloves per surgeon and per operation. We called this unit of analysis a surgeon‐operation to indicate that this was the perforation risk for one surgeon performing one operation.

Dealing with missing data

We contacted the authors of Meyer 1996, Nordkam 2005, Sullivan 2009, and Wilson 2008 to obtain information about the exact number of gloves and holes in gloves that had occurred during their study. Nordkam 2005 replied that a total of 812 gloves were used by two operators in the intervention arm and two operators in the control arm. The authors wrote that randomisation was achieved by computer generated numbers and that the observers of the outcome were blind to the type of needle used. We were told that the total number of holes was proportional to the number of perforations per operation. Given the total number of 56 holes, we used the proportion of perforations per operation to calculate the number of holes per study arm. Meyer 1996 replied they used a random number list which was not known to the surgeons. The surgeon called the study nurse while changing his gloves before closure of the laparotomy. For this reason they included only elective surgery in their study. The person who assessed the gloves identified the procedure by the patient´s name and operation book number. The surgeon´s opinion was evaluated separately. The authors used a closure technique in which the intestine was protected from needle stick injury by the non‐dominant hand of the surgeon who did the facial closure and not by instruments. This explains the preferential perforation of the left glove of a right‐handed surgeon and may in part explain the quantity of perforations.

For the ease of use of blunt needles for Nordkam 2005, we calculated standard deviations from P values according to the methods given by the Cochrane Handbook (Higgins 2011).

Assessment of heterogeneity

We defined clinically homogeneous studies as those with similar populations, interventions and outcomes measured at the same follow‐up point. We deemed all surgical staff to be similar enough to yield similar treatment effects. We also deemed all blunt and sharp needles to be similar enough to be combined. We did not combine different outcomes such as glove perforations and reported needle stick injuries. All glove perforations, even though measured in slightly different ways, are deemed sufficiently similar to be combined in a single comparison.

We assessed the amount of statistical heterogeneity in the meta‐analysis by means of the I² statistic. We used the criterion of an I² = 50%, as mentioned in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), to discern the presence or absence of considerable statistical heterogeneity between studies.

Assessment of reporting biases

We assessed publication bias with a funnel plot.

Data synthesis

We pooled studies with sufficient data, and judged to be clinically homogeneous, with RevMan software (RevMan 2011).

We intended to use a random‐effects model in the meta‐analysis when studies were statistically heterogeneous, but they weren't. We used a fixed‐effect model to pool study results.

For studies with multiple study arms that belonged to the same comparison, we intended to divide the number of events and participants equally over the study arms to prevent double counting of study participants in the meta‐analysis, but no such studies were found.

Finally, we used the GRADE approach to assess the quality of the evidence per comparison and per outcome as described in the Cochrane Handbook of Systematic Reviews (Higgins 2011). Starting from high quality, we reduced the quality of evidence by one or more levels if there was one or more limitations in any of the following domains: risk of bias, consistency, directness of the evidence, precision of the pooled estimate and the possibility of publication bias. We used the programme GRADEPro to generate a 'Summary of findings' table.

Subgroup analysis and investigation of heterogeneity

We investigated if the effects of interventions varied per type of operation involved when we compared abdominal closure and caesarean section versus vaginal repair and versus hip operations. This was however a subgroup analysis that we defined post‐hoc and was not specified in the protocol. We intended to compare surgeons versus their assistants but there were not sufficient data available. We also intended to compare different outcomes but they were measured in a similar way and no subgroup analysis was possible.

Sensitivity analysis

We analysed what the influence of studies with a high risk of bias was by re‐analysing the data only for studies with a low risk of bias.

Results

Description of studies

Results of the search

With the initial search strategy described in Appendix 1 and after removal of duplicates we had a total of 11,239 references. We selected 322 references for full‐text reading. This finally resulted in nine studies that fulfilled our inclusion criteria. In the additional search we found 211 references and one was read as full‐text. One extra study fulfilled our inclusion criteria. This brought the final number of included studies to 10 (Figure 1).

1.

Study flow diagram.

Included studies

The studies were carried out between 1998 and 2005. Seven studies were from before 2000.

The studies were located in the UK (Ablett 1998; Hartley 1996; Thomas 1995; Wright 1993), the US (Sullivan 2009; Wilson 2008), Ireland (Rice 1996), the Netherlands (Nordkam 2005 ), Italy (Mingoli 1996) and Germany (Meyer 1996).

Four studies focused on the risk during abdominal closure (Hartley 1996; Meyer 1996; Mingoli 1996; Nordkam 2005), two studies on vaginal repair (Ablett 1998; Wilson 2008), two studies on caesarean section (Sullivan 2009; Thomas 1995) and two on hip replacement operations (Rice 1996; Wright 1993).

In five studies (Ablett 1998; Hartley 1996; Rice 1996; Wilson 2008; Wright 1993) one surgeon per operation participated in the study, in three studies there was a surgeon and an assistant (Meyer 1996; Nordkam 2005; Sullivan 2009), in one study there was a surgeon and two assistants (Mingoli 1996), and in one study it was unclear how many participated (Thomas 1995). In some studies the number of participants was partly inferred from the number of gloves used.

All studies used blunt or tapered needles as the intervention and compared them to sharp or conventional suture needles.

All studies used perforations in gloves as the measure of outcome and the total number of perforations per operator per operation could be assessed for all but one study. Thomas 1995 reported only the number of perforated gloves per operation and not the number of holes. Only two studies used the combination of water and air tests to detect perforations (Ablett 1998; Hartley 1996); for one study we assumed that the combination was used because the study was reported by the same author that validated the test (Thomas 1995). All the other studies used a water test only (Table 2).

1. Control group perforation rate, ascertainment method, operation type.

Study	Ascertainment Method	Operation Type	Control Rate /operator /operation
Ablett 1998	air/water	vaginal repair	0.16
Hartley 1996	air/water	abdominal surgery	0.36
Mingoli 1996	water	abdominal surgery	0.34
Meyer 1996	water	abdominal surgery	1.75
Nordkam 2005	water	abdominal surgery	0.20
Rice 1996	water	hip replacement	0.31
Sullivan 2009	water	caesarean section	0.09
Thomas 1995	assumed air/water	caesarean section	0.50
Wilson 2008	water	vaginal repair	0.02
Wright 1993	water	hip replacement	1.00

The perforation rate varied greatly between studies, from two in every 100 surgeon‐operations to 175 in every 100 surgeon‐operations (Table 2). Abdominal operations seemed to provide the highest risk and vaginal repair the lowest. It should also be noted that the method of ascertaining perforations has been reported to be related to the number of perforations, with the combined method detecting 25% more perforations than the water test only (Smith 1988).

Excluded studies

Two studies were excluded because the intervention was a needle handling device and not a blunt needle (Bebbington 1996; Botet 1998). Two other studies were excluded because they were not randomised or controlled (Monz 1991; Tydeman 1995).

Risk of bias in included studies

Taken all together, the risk of bias in most studies was considered low given the technical nature of the intervention. There were three studies that scored as low risk of bias for four items (Figure 2; Figure 3).

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Six of the 10 studies used a proper method of randomisation and in four studies it was unclear. Allocation was concealed by the use of envelopes until before the operation or the closure of the abdominal fascia in two studies only (Ablett 1998; Nordkam 2005) and in one study by the involvement of a research nurse (Meyer 1996). It is difficult to conceive how influencing of the randomisation procedure could have led to different results. Since the needles were used in the same type of operation, for both intervention and control group, the risk of exposure was similar for both groups. The surgeons could not know in advance how big their risk of exposure would be in an operation and then, for example, mitigate the risk by choosing a blunt needle. Therefore we think that this has not had a great influence on the results.

Blinding

None of the studies could blind the surgeon for the type of needle used because the surgeons will always be aware what type of needle they are using. We indicated this as an unclear risk of bias. Theoretically, it is possible that the surgeon treats sharp needles more carefully than blunt ones and thus influences the risk of exposure.

Outcome assessment can be done blind to the type of needle used. This was only reported for five studies (Mingoli 1996;Meyer 1996; Nordkam 2005; Sullivan 2009; Wilson 2008). For the other studies it was unclear if observers were blinded.

Incomplete outcome data

Eight out of the 10 studies reported holes for all gloves used. Since the gloves were collected immediately after the operation this was fairly easy to achieve. In two studies we weren't sure if all gloves were reported because information was lacking.

Selective reporting

Eight out of 10 studies reported all outcomes that were described in the methods section.

Other potential sources of bias

In one study (Thomas 1995) so much essential information was lacking that we considered this as a serious source of bias.

Effects of interventions

See: Table 1

Blunted versus sharp suture needles

Outcome: number of glove perforations

Ten trials including 2961 surgeon‐operations compared the effect of blunt versus sharp needles on glove perforations and found a significant reduction of glove perforations, with a relative risk (RR) of glove perforation of 0.46 (95% confidence interval (CI) 0.38 to 0.54) (Analysis 1.1; Figure 4).

1.1. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 1 Glove perforation rate.

4.

Forest plot of comparison: 1 Blunt versus sharp suture needles, outcome: 1.1 Glove perforation rate.

Outcome: percutaneous injuries

Five studies reported the number of percutaneous injuries but in one study there were no injuries in the intervention and control groups (Hartley 1996). We could combine four studies in a meta‐analysis. The use of blunt needles decreased the risk of a needle stick injury by 69% (RR 0.31, 95% CI 0.14 to 0.68) (Analysis 1.2).

1.2. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 2 Needle Stick Injuries.

Outcome: surgeon satisfaction and needle performance

Data on needle performance could only be extracted for Nordkam 2005 and Meyer 1996. Nordkam 2005 showed that surgeons found the sharp needles 20% easier to use, on a VAS scale from 0 to 100, but Meyer 1996 reported that the blunt needles were easier to use even though clearly more force was needed (Analysis 1.2; Analysis 1.3; Analysis 1.4; Analysis 1.5).

1.3. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 3 Ease of use of Needle.

1.4. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 4 Ease of use of needle.

1.5. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 5 Force needed.

Rice 1996 reported that the surgeons had no difficulties with the use of the blunt needles. In Sullivan 2009, 92% of the surgeons reported that the blunt needles were acceptable but they were less satisfied with their use. Wilson 2008 reported that surgeons found the blunt needles significantly more difficult to use. In Wright 1993, the surgeons found the blunt needles slightly more difficult to use but they had minimal effect on their technique.

Sensitivity analysis

We made two subgroups of studies, with high and low risk of bias respectively, to evaluate the impact of risk of bias on our results. There was no difference between the two groups (Analysis 1.6).

1.6. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 6 Sensitivity to ROB.

We checked if the use of risk ratios for those studies that made it feasible would change the results but there was hardly a difference. We also checked if using the number of glove pairs as the nominator in the rates made a difference but it did not substantially change the result.

Subgroup analysis

Even though the effect of the intervention seems to be less in vaginal repair operations, it is difficult to draw conclusions given the small number of studies in each subgroup (Analysis 1.7).

1.7. Analysis.

Comparison 1 Blunt versus sharp suture needles, Outcome 7 Subgroups Type of Operation.

Publication bias

The funnel plot revealed that there was one small study with a very positive outcome. Otherwise the study outcomes are located in all quadrants of the plot (Figure 5).

5.

Funnel plot of comparison: 1 Blunt versus sharp suture needles, outcome: 1.1 Glove perforation rate.

Grade assessment

We found that, overall, there was no serious risk of bias in the included studies for the glove perforations outcome. The comparison was direct. There was no heterogeneity, the confidence interval was narrow, and there was no indication of publication bias. Therefore, we kept the level of evidence at high quality and did not downgrade it (Table 1).

For the needle stick injuries outcome, we also concluded that there was no serious risk of bias in the studies. The comparison was direct, there was no heterogeneity but the confidence intervals were much wider. There was no indication of publication bias. Because of the more imprecise results and the less objective nature of the outcome, we downgraded the evidence by one level to moderate quality.

Discussion

Summary of main results

There is high quality evidence that blunt suture needles lead to fewer perforations in surgical gloves than sharp needles. There is moderate quality evidence that blunt needles also reduce the number of needle stick injuries. The results were not influenced by the quality of the studies, albeit in abdominal closure the effect may be more noticeable than in vaginal repair operations.

Overall completeness and applicability of evidence

We performed a comprehensive search for studies that evaluated interventions to prevent needle stick injuries that was further supplemented with an additional search specifically for blunt needle evaluations. We are quite confident that we located all available studies. Most studies were done in the 1990s when surgeons became more aware of the risk of infection with HIV, but there were more recent studies too. The needles were used in a variety of operations and in a variety of countries. However, it was not always clear for which tissues the blunt needles were used. Some authors mentioned that they used them only for fascia and muscle but others were not explicit here. We found one study that described how blunt needles can be used for a subcuticular skin closure (Miller 1994) and because they do not penetrate skin they are actually easier to use for this type of skin closure. It is not clear how well blunt needles perform in other tissues. Knowing in which tissues they don't perform well and thus defining the limitations of blunt needles might help to improve their implementation. . The intervention is simple, easy to apply to many operative techniques and only slightly more costly than conventional needles. Even though there has been evidence of their effectiveness for almost two decades, blunt needles are still not in use where they should be. We found one fairly recent survey article that reported on the implementation of blunt needles in the US (Catanzarite 2007). It states that in 2005 in the US, 71% of perioperative nurses had never tried to use blunt needles. They also report that in a small US region that they surveyed, of 104 obstetricians and gynaecologists in 2006, 63% had never used blunt needles, 25% had tried them but did not use them anymore and only 12% currently used blunt needles. In another survey in France, blunt‐tipped needles were only available in 49% of the operating theatres and 55% of the surgeons said they had never used them (Tarantola 2006). Thus, there seems to be ample room for implementation of this preventive intervention. It is difficult to judge what makes surgeons resist their use. Catanzarite 2007 reports that some surgeons don't like the 'feel' of blunt needles because they move more slowly through the tissue and can produce a noticeable pop as they penetrate fascia. This seems to confirm the satisfaction ratings in the studies included in this review where surgeons say they are less easy to use but still acceptable. Habit and lack of experience are also mentioned but the authors state that the main factor in achieving a higher rate of use of blunt needles is to increase their availability in operating rooms and labour and delivery units.

Quality of the evidence

The overall quality of the evidence in the studies was good even though the reporting of the older studies could have been better. Given the technical nature of the intervention and the objective method of ascertaining that the gloves had holes, used in all studies, we believe that there is no reason to assume a high risk of bias. The newer studies that are better reported and thus score better with regard to risk of bias show similar results to the older studies. It was however disturbing to notice that the habit of testing for perforations has changed over time to water tests only, which were originally reported to be less sensitive than the air test and even less so than the combination of the two tests (Smith 1988; Smith 1990). Everyone who has once repaired a bicycle tyre might know from experience that the air test is very sensitive. Very disturbingly, to support the validity of the water test used most authors referred to an article in which the water test had been described but not validated and stated with great certainty that: "the gloves were tested with a previously validated method". Because all authors tested both study arms with the same test, we don't think that this has influenced the results of this review. A possible effect of detecting fewer perforations could be that the effectiveness of blunt needles has been underestimated.

We used two outcomes: perforations and self‐reported injuries. Perforations are a proxy outcome and they are only assessed as a dichotomous outcome: perforation or no perforation. It could be that blunt needles make worse holes than sharp needles and thus increase the exposure to blood or increase the risk of infection. However, there are laboratory studies that show that circular taper point needles make smaller holes and convey less blood into the glove than sharp cutting needles (Bennett 1994;Edlich 2003; Lefebvre 2008). As far as we are aware there are no similar experiments with blunt tapered circular needles. It seems however justified to assume that a similar effect would occur for blunt circular needles. The risk of infection with blunt needles has been reported in another Cochrane Review which found one study that showed no difference in wound infection with the use of blunt needles for abdominal closure (Anderson 2008).

Potential biases in the review process

We included studies regardless of language or status of publication, which reduces reporting bias. We did data extraction and risk of bias assessment in duplicate and resolved all differences by consensus.

Agreements and disagreements with other studies or reviews

This review confirms the beliefs of professional organisations, such as the American College of Surgeons, that blunt needles reduce the risk of exposure to blood and bodily fluids for surgeons and their assistants (ACS 2011; NIOSH 2008). We found one other non‐systematic review (Berguer 2005) that included only four of the studies that we included. Their conclusions were the same as ours.

Authors' conclusions

Implications for practice.

There is high quality evidence that the use of blunt suture needles appreciably reduces the risk of exposure to blood and bodily fluids for surgeons and their assistants over a range of operations.

Implications for research.

There is high quality evidence that the use of blunt needles is beneficial and it is unlikely that future research will change this conclusion.

Feedback

Lack of patient‐relevant outcomes, 12 January 2016

Summary

Name: Caroline Struthers

Email Address: caroline.struthers@csm.ox.ac.uk

Affiliation: EQUATOR Network

Comment: In the plain language summary there is absolutely no mention of any difference in patient outcomes between the blunt and sharp needle groups. I understand that was not the focus of the review, but to not mention it at all seems strange. It may be the advantage of using sharp needles in terms of patient outcomes is outweighed by the reduction in risk of needle stick injuries and HIV to the surgeons and theatre staff, but it would be good to have that quantified, or at least mentioned in the plain language summary.

I agree with the conflict of interest statement below:

I certify that I have no affiliations with or involvement in any organization or entity with a financial interest in the subject matter of my feedback.

Reply

We thank Ms Struthers for her important comment. We agree that patient‐relevant outcomes are an important element to consider when evaluating the effectiveness of new techniques for ensuring the occupational safety of surgical staff. Unfortunately none of the 10 studies included in our Cochrane review assessed patient‐relevant outcomes such as wound infection or failure of operations. However, another Cochrane review by Anderson 2008 evaluated different techniques for closing the abdominal wall and their effects on patient‐relevant outcomes. They included one study (Stafford 1998) that showed no statistical significant difference in wound infection at discharge when using blunt needles for abdominal closure, albeit the confidence interval was wide. This study was not included in our review because it did not measure needle stick injuries. Given the lack of evidence on patient‐relevant outcomes, we think this issue is worth further study.

We will take this issue better in to account when we update this review and we will then revise all relevant sections of text accordingly.

Contributors

Annika Saarto, Jos Verbeek, Jani Ruotsalainen

What's new

Date	Event	Description
15 March 2016	Feedback has been incorporated	We received a comment saying that our plain language summary made no reference to patient‐relevant outcomes. Unfortunately none of the 10 studies included in our Cochrane review assessed patient‐relevant outcomes such as wound infection or failure of operations. However, another Cochrane review evaluated different techniques for closing the abdominal wall and their effects on patient‐relevant outcomes. They included one study that showed no statistical significant difference in wound infection at discharge when using blunt needles for abdominal closure, albeit the confidence interval was wide. This study was not included in our review because it did not measure needle stick injuries. Given the lack of evidence on patient‐relevant outcomes, we think this issue is worth further study. We will take this issue better into account when we update this review. We will then revise all relevant sections of text accordingly.
8 May 2012	Amended	We changed risk ratios to rate ratios
27 February 2011	Amended	The original version of this protocol was published with the title: "Prevention of percutaneous injuries with risk of hepatitis B, hepatitis C, or other viral infections for health‐care workers". However, it turned out that the scope was far too wide and would result in an unmanageable amount of studies for one review. Therefore the decision was taken to split the protocol into four new ones. The other three new titles are: "Devices to prevent needle recapping for preventing percutaneous exposure injuries in health care personnel", "Education and training for preventing percutaneous exposure injuries in health care personnel" and "Extra gloves versus a single pair of gloves for preventing percutaneous exposure injuries in healthcare personnel".

Notes

The protocol for this review was first published as "Prevention of percutaneous injuries with risk of hepatitis B, hepatitis C, or other viral infections for healthcare workers" (Parantainen 2008). Our initial idea was to include all interventions used to prevent needle stick injuries. However, after the publication of the protocol it became apparent that very many studies would be eligible for inclusion. The decision was therefore made to split the protocol up into four new protocols, the first one being " Blunt versus sharp suture needles for preventing percutaneous exposure incidents in surgical staff" (Parantainen 2011). The remaining three protocols are titled: "Devices to prevent needle recapping for preventing percutaneous exposure injuries in health care personnel" (Lavoie 2012), "Education and training for preventing sharps injuries and splash exposures in healthcare workers" (Cheetham 2016) and "Extra gloves versus a single pair of gloves for preventing percutaneous exposure injuries in healthcare personnel" (Parantainen 2012).

The original protocol was hosted by the Hepato‐Biliary Group but due to the heavy involvement of Jos Verbeek and the Occupational Safety and Health Group, the new titles were registered under their aegis. The Hepato‐Biliary Group continues to be involved in an advisory capacity.

Appendices

Appendix 1. Search strategies

Database	Period of search	Search strategy
The Cochrane Hepato‐Biliary Group Controlled Trials Register	1996 to Sept 2010	('health care worker*' or 'health personnel' or 'HCWs' ) and ( 'virus disease*' or 'virus*' or 'viral infect*')
EMBASE	1974 to Sept 2010	#6 #5 AND [humans]/lim AND [embase]/lim #5 #3 AND #4 #4 [randomized controlled trial]/lim OR [controlled clinical trial]/lim OR random* OR 'double blind' OR 'single blind' OR (singl* OR doubl* OR trebl* OR tripl* AND (blind* OR mask*)) OR 'single blind procedure'/exp OR 'double blind procedure'/exp OR 'triple blind procedure'/exp OR placebo* OR 'controlled study'/exp OR 'cross sectional study'/exp OR 'crossover procedure'/exp OR 'latin square design'/exp OR 'follow up'/exp OR 'comparative study'/exp OR 'evaluation studies'/exp OR 'evaluation study' OR prospectiv* OR volunteer*] #3 #1 AND #2 #2 'health care personnel'/exp OR 'health care personnel' OR 'health care worker'/exp OR 'health care worker' OR 'health care workers' OR 'health care facilities and services'/exp OR 'medical profession'/exp OR 'nursing as a profession'/exp OR ('virus transmission'/exp AND 'patient'/exp AND professional) #1 'needlestick injury'/exp OR needlestick* OR 'needle stick'/exp OR 'sharp injury' OR 'sharp injuries' OR 'sharp medical' OR 'sharp instrument' OR 'sharp needle' OR 'sharp needles' OR sharps OR 'percutaneous exposure' OR 'percutaneous injury' OR 'percutaneous injuries' OR 'percutaneous trauma' OR 'stick injury' OR 'stick injuries' OR 'stab wound'/exp OR 'face injury'/de OR 'eye injury'/de OR 'arm injury'/de OR 'hand injury'/de OR 'needle'/exp OR (splash* AND ('blood'/exp OR blood OR secretion* OR fluid* OR 'body fluid'/exp OR 'body fluids'/exp))
Wiley InterScience: Cochrane Library Databases: CENTRAL and NHSEED	1993 to Sept 2010	#3 #1 AND #2 #2 EXP Needlestick Injuries (MeSH) OR needlestick* OR "needle stick OR "needle sticks" OR "percutaneous exposure" OR "percutaneous exposures" OR "percutaneous injury" OR "percutaneous injuries2 OR "stick injury" OR "stick injuries" OR Wounds, Stab (MeSH) OR Wounds, Penetrating (MeSH) OR Facial injuries (MeSH) OR EXP Eye Injuries, Penetrating (MeSH) OR Forearm Injuries (MeSH) OR EXP Hand Injuries (MeSH) OR [splash* AND blood OR secretion* OR fluid* OR EXP Body Fluids (MeSH)  OR EXP Bodily Secretions (MeSH)] #1 EXP Health Occupations (MeSH) OR EXP Health Personnel (MeSH) OR EXP Health Facilities (MeSH)  OR "health care worker"  OR "health care workers"  OR Disease Transmission, patient‐to‐Professional (MeSH)
Science Citation Index Expanded	1986 to 5 October 2010	#4 #1 AND #2 AND #3 #3 TS=(random* OR control* OR trial OR trials OR "single blind" OR "double blind" OR "triple blind" OR "latin square" OR placebo* OR comparative OR "follow up" OR prospectiv* OR "cross over" OR volunteer*) #2 TS=(needlestick* OR "needle stick" OR "needle sticks" OR "stick injury" OR "stick injuries"  OR "wound stab" OR "stab wound"  OR "penetrating wound" OR "penetrating wounds") OR TS=(sharp* AND ( injury OR injuries OR medical OR instrument*))  OR TS=(percutaneous AND (exposure OR exposures OR injury OR injuries)) OR TS=(injur* AND (facial OR eye OR eyes OR arm OR hand OR finger OR fingers)) OR TS=(splash* AND (blood OR secretion* OR fluid OR fluids)) OR TS="blood borne infection" #1 TS=("health care worker"  OR "health care workers" OR "health occupations" OR "health personnel"  OR physician* OR nurse* OR hospital* OR clinic OR clinics)
CINAHL	1982 to Sept 2010	#5 #3 AND #4 #4 "randomized controlled trial" or "clinical trials" or "clinical trial" or "random allocation" or "double blind". or "single blind" or ((singl* or doubl* or trebl* or tripl*) and (blind* or mask*)) or "latin square" or placebo# or random* or "research design" or "comparative study" or "comparative studies" or "evaluation study" or "evaluation studies" or "follow up study" or "follow up studies" or "prospective study" or "prospective studies" or "cross over study" or "cross over studies" or control* or prospective* or volunteer or (MH "Clinical Trials+") or (MH "Nonrandomized Trials") or (MH "Crossover Design") #3 #1 AND #2 #2 TX "needlestick injury" or needlestick# or "needle stick" or "needle sticks" or "sharp injury" or "sharp injuries" or "sharp medical device" or "sharp medical devices" or "sharp instrument" or "sharp instruments" or "sharp needle" or "sharp needles" or "percutaneous exposure" or "percutaneous exposures" or "percutaneous injury" or "percutaneous injuries" or "stick injury" or "stick injuries" or "wounds, stab" or "wounds, penetrating" or "facial injuries" or "eye injuries, penetrating" or "arm injuries" or "forearm injuries" or "hand injuries" or "finger injuries" or (splash# and (blood or secretion# or fluid#)) or ("occupational exposure" and ("body fluid" or  "body fluids" or blood)) #1 (MH "Health Occupations") OR health occupations   or (MH "Health Personnel+")  or (MH "Health Facilities+") OR health facilities  or TX "health care worker"  or TX "health care workers"  or (MH "Personnel, Health Facility+") or (MH "Occupational Health Services+")  or (MH "Occupational Hazards+")  or (MH "Occupational Exposure")  or TX "health care personnel"   or  (MH "Health Personnel+") or (MH "HIV Infections+")
OSH UPDATE (NIOSHTIC‐2 and CISDOC)	NIOSHTIC‐2: 1900 to Sept 2010 CISDOC: 1987 to Sept 2010	#15 #13 AND #14 #14 PY{2007} OR PY{2008} OR PY{2009} #13 #7 AND #12 #12 #8 OR #11 #11 #9 AND #10 #10 GW{blind* OR mask*} #9 GW{singl* OR doubl* OR tripl* OR trebl*} #8 GW{random* OR control* OR trial OR trials OR comparativ* OR evaluation* OR "latin square" OR placebo OR "follow up" OR prospectiv* OR "cross over" OR volunteer*} #7 #1 AND #6  #6 #2 OR #5 #5 #3 AND #4 #4 GW{splash*} #3 GW{blood OR fluid* OR secretion*} #2 GW{"sharp medical" OR "sharp instrument" OR "sharp instruments" OR needlestick* OR "needle stick" OR "needle sticks" OR "sharp injury" OR "sharp injuries" OR "stab wound" OR "stab wounds" OR "wound penetrating" OR "stick injury" OR "stick injuries" OR "percutaneous injury" OR "percutaneous injuries" OR "percutaneous exposure" OR "percutaneous exposures" OR "sharp needle" OR "sharp needles"} #1 GW{nurse OR nurses OR physician OR physicians OR hospital* OR "health occupation" OR "health occupations" OR "health personnel" OR "health care personnel" OR "health care worker" OR "health care workers" OR "health worker" OR "health workers"}
MEDLINE in PubMed	from 1950 to Sept 2010	#5 Search #1 AND #2 AND (#3 OR #4) #4 Search effect*[tw] OR control[tw] OR controls*[tw] OR controla*[tw] OR controle*[tw] OR controli*[tw] OR controll*[tw] OR control'*[tw] OR evaluation*[tw] OR program*[tw] #3 ("Randomized Controlled Trial"[pt] OR "Controlled Clinical Trial"[pt] OR "Randomized Controlled Trials as Topic"[mh] OR "Random Allocation"[mh] OR "Double‐Blind Method"[mh] OR "Single‐Blind Method"[mh] OR "Clinical Trial"[pt] OR "Clinical Trials as Topic"[mh] OR "clinical trial"[tw] OR ((singl*[tw] OR doubl*[tw] OR trebl*[tw] OR tripl*[tw]) AND (mask*[tw] OR blind*[tw])) OR "latin square"[tw] OR Placebos[mh] OR placebo*[tw] OR random*[tw] OR "Research Design"[mh:noexp] OR "Comparative Study"[pt] OR "Evaluation Studies as Topic"[mh] OR "Follow‐up Studies"[mh] OR "Prospective Studies"[mh] OR "Cross‐over Studies"[mh] OR control[tw] OR controls*[tw] OR controla*[tw] OR controle*[tw] OR controli*[tw] OR controll*[tw] OR control'*[tw] OR prospectiv*[tw] OR volunteer*[tw]) NOT (Animals[mh] NOT Humans[mh) #2 "Needlestick injuries"[mh] OR needlestick*[tw] OR "needle stick"[tw] or "needle sticks"[tw] OR "sharp injury"[tw] OR "sharp injuries"[tw] OR sharps[tw] OR "sharp medical device"[tw] OR "sharp medical devices"[tw] OR "sharp instrument"[tw] OR "sharp instruments"[tw] OR "sharp medical instrument"[tw] OR "sharp medical instruments"[tw] OR "sharp needle"[tw] OR "sharp needles"[tw] OR "percutaneous exposure"[tw] OR "percutaneous exposures"[tw] OR "percutaneous injury"[tw] OR "percutaneous injuries"[tw] OR "stick injury"[tw] OR "stick injuries"[tw] OR "Wounds, Stab"[mh:noexp] OR "Wounds, Penetrating"[mh:noexp] OR "Facial injuries"[mh:noexp] OR "Eye Injuries, Penetrating"[mh] OR "Arm Injuries"[mh:noexp] OR "Forearm Injuries"[mh:noexp] OR "Hand Injuries"[mh] OR (splash* AND (blood[tw] or secretion*[tw] OR fluid*[tw] OR "Body Fluids"[mh])) #1 "Health Occupations"[mh] OR "Health Personnel"[mh] OR "Health Facilities"[mh] OR "health care worker"[tw] OR "health care workers"[tw] OR "Infectious Disease Transmission, Patient‐to‐Professional"[mh]
PsycINFO (OvidSP)	1967 to Sept 2010	#5 limit 4 to all journals #4 #1 AND #2 AND #3 #3 random* OR control* OR trial OR trials OR comparativ* OR evaluation* OR ((singl* OR doubl* OR tripl* OR trebl*) AND (blind* OR mask*)) OR "latin square" OR placebo* OR "follow up" OR prospectiv* OR "cross over" OR volunteer*    #2 (splash* AND (blood OR secretion* OR fluid OR fluids)) OR ("eye injuries" AND penetrating) OR (wound* AND (stab OR penetrating)) OR "percutaneous exposure" OR "percutaneous exposures" OR "percutaneous injury" OR "percutaneous injuries" OR "stick injury" OR "stick injuries" OR "sharp injury" OR "sharp injuries" OR "sharp medical" OR "sharp instrument" OR "sharp instruments" OR "sharp needle" OR "sharp needles" OR needlestick* OR "needle stick" OR "needle sticks" #1 (nursing or nurse or nurses or physician or physicians or "health care personnel" or "health personnel" or "health care worker" or "health care workers" or "Clinicians*" or "Dentist*" or "Health‐Personnel" or "Medical Personnel" or "Military‐Medical‐Personnel" or "Nurses*" or "Physician*" or "Psychiatric‐Hospital‐Staff*" or "medical students" or "hospitals" or "occupational exposure" or "occupational exposures").mp. [mp=title, abstract, heading word, table of contents, key concepts]
LILACS	Sept 2010	"Health Occupations" or "Health Personnel" OR "Health Facilities" OR "health care worker" OR "health care workers" OR "Disease Transmission, Patient‐to‐Professional" OR "INJURIES" or "WOUNDS AND INJURIES/PC" or "accidents, OCCUPATIONAL" or "injuries, poisonings, and OCCUPATIONAL diseases" or "OCCUPATIONAL exposure" or "OCCUPATIONAL health policy" or "OCCUPATIONAL risks" OR "INJURIES" or "WOUNDS AND INJURIES/PC" or "accidents, OCCUPATIONAL" or "injuries, poisonings, and OCCUPATIONAL diseases" or "OCCUPATIONAL exposure" or "OCCUPATIONAL health policy" or "OCCUPATIONAL risks" [Descritor de assunto] and "CLINICAL TRIAL" OR "CLINICAL TRIAL, PHASE I" OR "CLINICAL TRIAL, PHASE II" OR "CLINICAL TRIAL, PHASE III" OR "CLINICAL TRIAL, PHASE IV" OR "COMPARATIVE STUDY" OR "CONTROLLED CLINICAL TRIAL" OR "EVALUATION STUDIES" OR "META‐ANALYSIS" OR "MULTICENTER STUDY" OR "RANDOMIZED CONTROLLED TRIAL" OR "REVIEW" [Tipo de publicação] and not "ANIMALS" or "HUMANS" [Palavras]

Appendix 2. Updated search 27 April 2011

PubMed

(taper* OR blunt* OR protective) AND Needle* AND ("Randomized Controlled Trial"[pt] OR "Controlled Clinical Trial"[pt] OR "Randomized Controlled Trials as Topic"[mh] OR "Random Allocation"[mh] OR "Double‐Blind Method"[mh] OR "Single‐Blind Method"[mh] OR "Clinical Trial"[pt] OR "Clinical Trials as Topic"[mh] OR "clinical trial"[tw] OR ((singl*[tw] OR doubl*[tw] OR trebl*[tw] OR tripl*[tw]) AND (mask*[tw] OR blind*[tw])) OR "latin square"[tw] OR Placebos[mh] OR placebo*[tw] OR random*[tw] OR "Research Design"[mh:noexp] OR "Comparative Study"[pt] OR "Evaluation Studies as Topic"[mh] OR "Follow‐up Studies"[mh] OR "Prospective Studies"[mh] OR "Cross‐over Studies"[mh] OR control[tw] OR controls*[tw] OR controla*[tw] OR controle*[tw] OR controli*[tw] OR controll*[tw] OR control'*[tw] OR prospectiv*[tw] OR volunteer*[tw]) NOT (Animals[mh] NOT Humans[mh) N=68 references

EMBASE

(taper* OR blunt* OR protective) AND needle* AND ([randomized controlled trial]/lim OR [controlled clinical trial]/lim OR random* OR 'double blind' OR 'single blind' OR (singl* OR doubl* OR trebl* OR tripl* AND (blind* OR mask*)) OR 'single blind procedure'/exp OR 'double blind procedure'/exp OR 'triple blind procedure'/exp OR placebo*) N=211 references

Data and analyses

Comparison 1. Blunt versus sharp suture needles.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Glove perforation rate	10		Rate Ratio (Fixed, 95% CI)	0.46 [0.38, 0.54]
2 Needle Stick Injuries	5	1017	Risk Ratio (M‐H, Fixed, 95% CI)	0.31 [0.14, 0.68]
3 Ease of use of Needle	1	400	Mean Difference (IV, Fixed, 95% CI)	‐18.0 [‐28.64, ‐7.36]
4 Ease of use of needle	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	1.14 [1.01, 1.28]
5 Force needed	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.31 [0.19, 0.51]
6 Sensitivity to ROB	10		Rate Ratio (Fixed, 95% CI)	0.46 [0.38, 0.54]
6.1 High Risk of Bias	7		Rate Ratio (Fixed, 95% CI)	0.45 [0.38, 0.55]
6.2 Low Risk of Bias	3		Rate Ratio (Fixed, 95% CI)	0.46 [0.29, 0.72]
7 Subgroups Type of Operation	10		Risk Ratio (Fixed, 95% CI)	0.46 [0.38, 0.54]
7.1 Abdominal Closure and Cesarean Section	6		Risk Ratio (Fixed, 95% CI)	0.45 [0.37, 0.54]
7.2 Vaginal Repair	2		Risk Ratio (Fixed, 95% CI)	0.59 [0.30, 1.20]
7.3 Hip Replacement	2		Risk Ratio (Fixed, 95% CI)	0.43 [0.24, 0.76]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Ablett 1998.

Methods	Study design: randomised controlled trial. Object of randomisation: patients.
Participants	UK. HCWs performing suturing of perineal tears and episiotomies including perineal skin. Intervention group n = 104 surgeon‐operations. Control group n = 91 surgeon operations. Number observed: 195 pairs of gloves.
Interventions	Suturing with blunt tipped needles. The control group used sharp needles.
Outcomes	Outcome: The number of glove perforations and reported needlestick injuries. Perforation detection: The gloves were filled with air and then immersed in water and perforations were noted as air bubbles.The gloves were thereafter filled with water and perforations were noted as jets of water.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	randomised allocation
Allocation concealment (selection bias)	Low risk	into unmarked envelopes
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	into unmarked envelopes; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	not reported
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	not reported
Selective reporting (reporting bias)	Low risk	all outcomes reported

Hartley 1996.

Methods	Study design: randomised controlled trial. Objects randomised: patients.
Participants	UK. Consultants or registrars performing mass closure of the abdomen: fascial and muscle layers. Number studied: 85 patients. Intervention group n = 46 surgeon‐operations, control group n = 39 surgeon‐operations, 85 pairs of gloves observed. Gloves changed before closure commenced.
Interventions	Use of suture needles with blunted end. Control group used conventional sharp pointed suture needles.
Outcomes	Number of perforated glove pairs per total number of procedures. Perforation detection: firstly the gloves were filled with air, immersed in water and perforations were noted as air bubbles. Secondly, the gloves were filled with water and perforations were noted as jet of water. For the second half of the study additional testing was performed with a specifically designed electrical testing device.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	randomisation by random number generator
Allocation concealment (selection bias)	Unclear risk	not reported
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	not reported; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	no attrition
Selective reporting (reporting bias)	Low risk	all outcomes reported

Meyer 1996.

Methods	Study design: randomised controlled trial according to operation date and time. Object of randomisation: operations
Participants	Germany. Attending surgeon and assistant during abdominal closure: fascia and muscle. Number studied 100 operations (400 gloves). Intervention group n = 98 surgeon‐operations (196 gloves) Control group n = 102 surgeon‐operations (204 gloves) Gloves changed before closure.
Interventions	Use of blunt needles to close the abdominal fascia. Control group used sharp needles
Outcomes	Outcome 1. Number of perforations in gloves. 2. Force needed to use the needle 3. Ease of use of the needle. Gloves were filled with water and perforations were noted as jets of water
Notes	Information from author:The surgeon called the study nurse while changing his gloves before closure of the laparotomy. Therefore, only elective surgery has been included in this study.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation was done according to operation date and time according to a random number list.
Allocation concealment (selection bias)	Low risk	Randomisation was done according to operation date and time
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not reported; not possible
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The person, who assessed the globes, identified the procedure by participant´s name + OP‐book number. Surgeon´s opinion was evaluated separately. (information from author)
Incomplete outcome data (attrition bias) All outcomes	Low risk	All gloves reported
Selective reporting (reporting bias)	Low risk	All outcomes reported

Mingoli 1996.

Methods	Study design: randomised controlled trial. Object of randomisation: procedures.
Participants	Italy. Attending surgeons, 3rd to 4th year residents and medical students performing abdominal procedures: muscle and fascia. Number studied: 100 operations (1560 gloves). Intervention group n = 300 surgeon operations (780 gloves). Control group n = 300 surgeon‐operations (780 gloves). Gloves changed before closure.
Interventions	Use of blunt needles in closure of the abdominal fascia. Control group used sharp needles.
Outcomes	Outcome: 1. Number of needlestick injuries per total number of operations; 2. Number of perforations per total number of gloves. Perforation detection: The gloves were filled with water and perforations were noted as jets of water.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	by random number allocation
Allocation concealment (selection bias)	Unclear risk	not reported
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	not reported; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Low risk	observers assessed outcome
Incomplete outcome data (attrition bias) All outcomes	Low risk	all gloves tested
Selective reporting (reporting bias)	Low risk	all outcomes reported

Nordkam 2005.

Methods	Study design: randomised controlled trial. Object of randomisation: patients.
Participants	Netherlands. Surgeons and assistants performing laparotomies, abdominal closure: fascia. Number studied: 200 procedures, 400 operators and 406 pairs of gloves. Intervention group n = 200 surgeon‐operations. Control group n = 100 surgeon‐operations. Gloves changed before skin closure.
Interventions	Use of blunt‐tapered needles. Control group used sharp needles.
Outcomes	Outcome: Number of procedures with one or more glove perforations per total number of procedures. Outcome calculated: holes in gloves per operator per operation Perforation detection: The gloves were filled with water and perforations were noted as jets of water. Secondary outcome: Ease of use of the needles.
Notes	We received additional information from the authors
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	randomisation by envelope; authors wrote that randomisation was based on a computer generated random number
Allocation concealment (selection bias)	Low risk	"just before closure of the fascia"; randomisation by envelope
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	team was aware of type of needle used; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Low risk	not reported; authors wrote that observer was blind to which needle was used
Incomplete outcome data (attrition bias) All outcomes	Low risk	all gloves retrieved
Selective reporting (reporting bias)	Low risk	all outcomes reported

Rice 1996.

Methods	Study design: randomised controlled trial. Object of randomisation: operations.
Participants	Ireland. Surgeons performing deep wound closure in primary total hip replacement surgery: fascia, fat and muscle not skin. Number studied: 68 operations. Intervention group n = 34 surgeon‐operations (number of gloves not reported). Control group n = 34 surgeon operations (128 gloves). Fresh outer gloves put on before wound closure and changed again before skin closure.
Interventions	Use of taper pointed suture needle. Control group used standard pointed needles.
Outcomes	Outcome: 1. Number of perforations per total number of gloves; 2. Number of glove perforations per total number of operations. Perforation detection: The gloves were filled with water and perforations were noted as jets of water.
Notes	We assumed that the number of glove pairs in experimental group was the same as the number of procedures, and used outer glove perforations as outcome.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	operation randomly selected
Allocation concealment (selection bias)	Unclear risk	not reported; blinding to type of needle is impossible for surgeons
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	not reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	4/68 pairs of gloves missing
Selective reporting (reporting bias)	Low risk	all outcomes reported

Sullivan 2009.

Methods	Study design: randomised controlled trial. Object of randomisation: operations.
Participants	USA. Surgeons and assistants performing caesarean sections. Closure of all anatomical layers. Number studied: 194 operations. Intervention group n = 204 surgeon‐operations. Control group n = 204 surgeon‐operations. All gloves collected.
Interventions	Use of blunt needles. The control group used sharp needles.
Outcomes	Outcome: number of glove perforations per operation. Perforation detection: the gloves were filled with water and perforations were noted as jets of water.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	computerised random sequence generator
Allocation concealment (selection bias)	Unclear risk	surgeon nor assistant blinded to randomisation outcome
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	surgeon nor assistant blinded to randomisation outcome; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Low risk	the (testing) author was blind to which needle had been used
Incomplete outcome data (attrition bias) All outcomes	Low risk	100% completion of surveys
Selective reporting (reporting bias)	Low risk	all outcomes reported

Thomas 1995.

Methods	Study design: randomised controlled study. Object of randomisation: procedures.
Participants	UK. Surgeons performing caesarean sections, closing the uterine wound. Number studied: 80 procedures. Sizes of intervention or control group not reported but assumed to be 40 and 40 surgeon‐operations. Glove change not reported.
Interventions	Use of blunt‐tipped needles. Control group used sharp‐tipped needles.
Outcomes	Outcome: percentage of perforated gloves. Detection method not reported but we assumed that this was the combined air and water method developed by one of the authors.
Notes	We used the number of perforated gloves per total number of procedures as outcome.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	a random allocated study
Allocation concealment (selection bias)	Unclear risk	not reported
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	not reported; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	not reported
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	not reported
Selective reporting (reporting bias)	Unclear risk	no description in methods
Other bias	High risk	The study design made 160 patients desirable but due to surgeons preference at that time only 80 patients were enrolled

Wilson 2008.

Methods	Study design: randomised controlled trial. Object of randomisation: patients.
Participants	USA. Physicians performing obstetrical laceration repairs, needle used for whole procedure. Number studied: 438 procedures. Intervention group n = 217 surgeon‐operations. Control group n = 221 surgeon operations. All gloves collected.
Interventions	Use of blunt needles. Control group used sharp needles.
Outcomes	Outcome: number of glove perforations per total number of operations (also in RevMan analyses) but all surgeons were single gloved. Perforation detection: The gloves were filled with water and perforations were noted as jets of water. Secondary outcome: user satisfaction.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	computer random sequence generator
Allocation concealment (selection bias)	Unclear risk	surgeons not blinded to randomisation outcome
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	surgeons not blinded to randomisation outcome; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Low risk	the investigator was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	all pairs of gloves were collected
Selective reporting (reporting bias)	Low risk	all outcomes reported

Wright 1993.

Methods	Study design: randomised controlled trial. Object of randomisation: operations.
Participants	UK. 17 surgeons performing standard wound closure after hip arthroplasties, skin closed with staple machine. Number studied: 69 operations. Surgeons were double gloved. We used only holes in outer gloves for the outcome. Intervention group n = 38 surgeon‐operations. Control group n = 31 surgeon‐operations. Outer pair of gloves changed before insertion of prosthetic components and before wound closure.
Interventions	Use of blunt taper point suture needles. Control group used standard cutting needles.
Outcomes	Outcome: number of outer glove perforations per total number of outer gloves. Perforation detection: The gloves were filled with water and perforations were noted as jets of water. Secondary outcome: Self‐reports of needlestick injuries, ease of use of the needle.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	allocated preoperatively by a random number
Allocation concealment (selection bias)	Unclear risk	not reported
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	not reported; blinding to type of needle is impossible for surgeons
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	138 gloves from 69 operations reported
Selective reporting (reporting bias)	Unclear risk	all outcomes reported

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Bebbington 1996	Not a blunt needle but a device
Botet 1998	Not a blunt needle but a device
Monz 1991	Not a randomised study
Tydeman 1995	Not a randomised study

Differences between protocol and review

We performed a subgroup analysis of types of operations that was not stated in the protocol.

Contributions of authors

Annika Parantainen is the co‐ordinator of the review work. She participated in the study selection, the data extraction and the writing of the review.
 Marie‐Claude Lavoie and Manisha Pahwa participated in the study selection and the data extraction.
 Jos Verbeek participated in the study selection, the data extraction, offered methodological and statistical support and wrote the text of the review.

Sources of support

Internal sources

• Occupational Health and Safety Review Group, Finland.

External sources

• FinOHTA, Finland.

Declarations of interest

There are no conflicts of interests.

Edited (no change to conclusions), comment added to review