Article Text
PDF
Abstract
Background Urethral swabs are uncomfortable due to the nature of the mucosa and may be a reason for non-attendance of men at sexually transmitted infection (STI) clinics. This randomised controlled trial describes the extent of discomfort associated with direct urethral sampling, and determines whether this varies by the type of swab used.
Methods Male patients over the age of 16 years having swab tests were invited to participate and were randomly assigned to receive either a gonorrhoea dacron-tipped swab, a chlamydia rayon-tipped swab or a plastic 10 μl urethral loop first for urethral sampling followed by the others sequentially. Discomfort was measured using a 0–100 mm visual analogue scale (VAS).
Findings 129 men having urethral swabs carried out as part of their screening tests for STI were invited to participate in the study and 121 were recruited. The median pain scores (IQR) regardless of sampling method, before and after the urethral sampling were, first 0 mm (0–0) and 50 mm (22–71) (p<0.001), second 9 mm (0–28) and 59.5 mm (38.3–78) (p<0.001) and third 10 mm (0–31) and 58 mm (29.3–80) (p<0.001). Direct urethral sampling was associated with a median pain score of 60.5 mm using a rayon swab, 52 mm using a dacron swab and 25.5 mm using a plastic loop.
Interpretation Direct urethral sampling is associated with discomfort and/or pain in men, which was significantly greater with a swab than a plastic loop. Urine should therefore be the specimen type of choice. When direct urethral sampling is indicated a loop is preferable to a urethral swab.
International Standard Randomised Controlled Trial Number Register number ISRCTN50938901.
- Chlamydia trachomatis
- genital specimens
- men
- Neisseria gonorrhoeae
- pain
- randomised controlled crossover trial
- sexually transmitted infections
- trials
- urethral swab
Statistics from Altmetric.com
- Chlamydia trachomatis
- genital specimens
- men
- Neisseria gonorrhoeae
- pain
- randomised controlled crossover trial
- sexually transmitted infections
- trials
- urethral swab
Direct urethral sampling is uncomfortable due to the nature of the mucosa and it has been postulated that a painful urethral swab may be a reason for non-attendance of men at sexually transmitted infection (STI) clinics.1 A search of the literature has not revealed any significant data on the intensity or nature of the pain felt during urethral swabbing. There is also a lack of data on inter-patient variability of pain and the effect of different healthcare workers (HCW) on the discomfort experienced with a urethral swab. Anecdotal evidence suggests that a plastic loop used for urethral sampling is less painful than a cotton wool or synthetic fibre-tipped swab, but we have been unable to determine from the literature if there is a difference in the discomfort experienced when different types of swabs are used to take a urethral specimen.
There have been studies investigating the effect of instilling 2% lignocaine hydrochloride gel before cystoscopies.2 3 These studies suggest that the lower the temperature of the gel, the better the anaesthetic effect. Another study has, however, also suggested that aqueous gel is better than lignocaine gel for pain relief before cystoscopies.4 It is not possible, however, to try out these strategies before a urethral swab as the gels usually contain an antibacterial agent and it precludes the use of the swab to provide a smear specimen for staining and microscopy.
This study was designed to describe the extent of discomfort associated with direct urethral sampling, and to determine whether this varies by the type of swab used.
Methods
Trial design
This was a randomised controlled crossover trial conducted at a single sexual health clinic in Derby, UK.
Population studied
Male patients over the age of 16 years attending the sexual health clinic during the study period and having swab tests were invited to participate in the study. Patients who had passed urine within 2 h of the swab, those who had taken analgesics, antidepressants or anti-epileptic drugs in the previous 24 h were excluded from the study.
Randomisation method
A block randomisation of the six allocated sequences of the three types of swab concealed in sealed opaque numbered envelopes was constructed by the medical statistician. The envelopes were opened sequentially by a member of the research team, only after the participant's name and clinic number were written on the appropriate envelope.
Swabbing protocol
Following informed signed consent by the principal investigator, participants were randomly assigned to receive one of the following types of urethral sampling methods first: a gonorrhoea dacron-tipped swab (manufactured by Copan, Italy); a chlamydia rayon-tipped swab (manufactured by Remel, Lenexa, USA) or a plastic 10 μl urethral loop for a urethral slide. All patients had the other two types of urethral sampling methods carried out in the order prespecified by the randomised envelopes. The same HCW performed the three urethral samples for each participant.
Data collection
Participants in the study were given a short demographics and symptom questionnaire to fill in before the first swab was taken for investigations. Data were collected on the frequency of micturition as defined by passing urine more than eight times during the day in the week before the test, the presence of a urethral discharge in the week before the test, and the presence of any pain on urination in the week before the test. Data were also collected on the presence and severity of urethral symptoms using a visual analogue scale (VAS). This was obtained before the swab as well as the identity of the HCW taking the swab.
The discomfort felt during the urethral sampling procedure when a loop or swab was inserted 2 cm into the urethra was assessed using the 0–100 mm VAS, which was the primary outcome.
The presence and concentration of epithelial cells and polymorphonuclear cells in the Gram-stained urethral smear under a (×1000) microscopic field (averaged over five fields with the greatest concentration of cells) was collected. This urethral smear was obtained from the plastic loop.
Data analysis
Sample size estimation
The smallest clinically important difference in the discomfort felt during direct urethral sampling using VAS scores was set at 20 mm.5 6 An initial power calculation was carried out based on the assumption of a SD of 40 mm, but an interim analysis of the data after the first 90 patients had been recruited showed that the highest observed SD of VAS scores in the study was only 26 mm. Therefore a power calculation based on this SD showed the study needed to recruit 37 patients per group to detect the clinical difference of 20 mm as significant (with a power of 90% and significance level of 5%).
Analytical methods
Collected data were entered into a database and analysis was carried out using the statistical package STATA version 11.0. The distribution of the continuous variables was checked using the Kolmogorov–Smirnov test and their histograms. The results of the skewed variables were presented as median (IQR) or otherwise as mean (±SD). The Kruskal–Wallis test was used for comparing the pain scores between the three different types of urethral sampling method and the Mann–Whitney U test was used for comparing pain scores between two types. Spearman correlations were used for correlating the VAS scores with the continuous variables. The pain scores before and after urethral sampling for each partipant were compared using paired analysis, Wilcoxon signed rank test.
Analysis of variance for a crossover study was used for testing the period, carryover, treatment and sequence effect on the change of pain scores, pre and post-urethral sampling.
A mixed-effects model with random effects at individual level (using maximum restricted likelihood) was used to explore the effect of the urethral sampling method, the presence of epithelial cells on the slide, the order of the urethral sampling (ie, if the sample was taken first, second or third), the sequence of urethral sampling, if the patient had previous experience of urethral sampling, and the interaction of the type and order of urethral sampling on the change of pain scores, before and after each swab. Only the significant factors from the univariate analysis were included in the mixed-effects model. The scatter plot of the standardised deviance residuals against the predicted values and the Q–Q plot for normality were used for testing the good fit of each model. The results are presented as a mean difference (95% CI) on increase in pain scores.
Results
Recruitment
From June 2008 to May 2009 129 men having urethral sampling carried out as part of their screening tests for STI were invited to participate in the study and 121 were recruited. Eight men declined to participate. The participant flow diagram is shown in flow diagram 1 in the online appendix.
Five HCW collected the urethral samples (three doctors and two nurses) from 62 (51.2%), eight (6.6%), 37 (30.6%), four (3.3%) and 10 (8.3%) participants, respectively.
Participant characteristics
Participant characteristics are documented in table 1.
Participant characteristics
The urethral sampling order for the first, second and third samples taken (number of samples taken first, second and third) were rayon (40, 41, 40), dacron (41, 40, 40) and plastic loop (40, 40, 41).
Pain associated with direct urethral sampling
The paired analysis of the pain scores show significant differences between the pain scores regardless of the urethral sampling method, as the median (IQR) scores before and after the first urethral sampling were 0 mm (0–0) and 50 mm (22–71) (p<0.001), before and after the second urethral sampling were 9 mm (0–28) and 59.5 mm (38.3–78) (p<0.001) and before and after the third urethral sampling were 10 mm (0–31) and 58 mm (29.3–80) (p<0.001).
Pain score by urethral sampling method
The plastic loop had significantly lower pain scores compared with rayon and dacron after the first, second and third urethral sampling. The dacron swab was also significantly less painful than the rayon swab after the first and third urethral sampling (table 2).
Median (IQR) pain scores post urethral sampling
Univariate analysis of factors affecting pain scores
Urinary symptoms
Participants with dysuria in the week before the test had significantly higher pain scores on assessment of severity of pain on urination than those without (15.5 (2.2–38.2) vs 0 (0–0), p<0.001) and had higher pre-first swab pain scores than those without (0 (0–17.5) vs 0 (0–0), p<0.0001). Dysuria did not affect other pain scores. Participants with frequency of micturition had higher pain scores on assessment of severity of pain on urination than those without (5 (0–33) vs 0 (0–0), p<0.001). Participants with a urethral discharge in the week before the test had higher pain scores with micturition (0 (0–23.5) vs 0 (0–0), p=0.005), higher pre-first urethral sampling pain scores (0 (0–21.5) vs 0 (0–0), p=0.02) and lower pre-third urethral sampling pain scores than those without urethral discharge (1 (0–19.5) vs 13 (0–33), p=0.02).
Findings on urethral slide
The presence of pus cells on the urethral slide or the concentration of pus cells did not affect the pre or post-urethral sampling pain scores; however, the presence of epithelial cells on the urethral slide was associated with a lower pain score after the second urethral sampling (52 (28.7–75.2) vs 75 (52–89.2), p=0.002).
Patient characteristics
The older participants were more likely to be married (p=0.0082), to have had urethral sampling previously (p=0.0024) and to have had an STI previously (p=0.0055). Participants identifying themselves as homosexual men had higher pre-first urethral sampling pain scores (17 (0–64)) than heterosexual (0 (0–0)) and bisexual individuals (0 (0–0), p=0.0296). Participants who had previously had swab tests had lower post-second urethral sampling pain scores (52 (24–78) vs 69 (56–78.5), p=0.02).
Healthcare worker
The identity of the HCW had no significant effect on the pain scores.
Crossover analysis
The effect of the urethral sampling method—dacron, rayon or loop—was significant in the increase in the pain scores before and after each urethral sampling (table 3). Moreover, an inter-patient variability of the pain scores was significantly observed. Due to this significant result mixed-effects regression analysis was performed taking into account the inter-patient variability. The effect of the sequence and order of urethral sampling as well as the carryover effect, from one urethral sampling method to another, was not significant.
Analysis of variance for a crossover study
Mixed-effects model on change of pain scores
Following the mixed-effects regression analysis, whereas the order and the sequence of urethral sampling did not have any significant effect on the increase in pain scores, the method of urethral sampling had a significant effect (table 4). The rayon and dacron swabs had significantly higher increases in pain scores compared with the plastic loop. The rayon swab was the most painful of all three types of urethral sampling method. It increased the pain scores by 31.2 mm more than the loop when they both were taken first, by 23.3 mm more when the rayon swab was taken second and by 20.2 mm when the rayon swab was taken third. On the other hand, the dacron swab did not increase the pain scores significantly more than the loop, when taken first or third but did so when it was taken second, by 17.9 mm more than the loop.
Mixed-effects model on increase in pain scores from pre-swab to post-swab: multivariate analysis
Participants with epithelial cells on the urethral slide were not more likely to experience an increase in pain after each swab than those without epithelial cells. Participants who had experienced urethral sampling in the past were more likely to have a significantly smaller increase in pain than those who did not have previous experience. Results of the multivariate analysis are presented in table 4.
Discussion
This study was designed to provide information on the level of discomfort, if any, experienced when a direct urethral sample is taken from men. Obtaining a direct urethral sample was associated with a median pain score of approximately 50 mm on a 100 mm pain scale, regardless of the urethral sampling method. Statistical analysis, however, showed that there was no significant carryover effect from one swab to another and that the pain was from inter-subject variability and the effect of the urethral sampling method. Direct urethral sampling thus induces pain/discomfort, of a magnitude correlating with moderate pain (>30 mm on a 100-point scale) or severe pain (>54 mm on a 100-point scale).7 There is a difference in the participants' rating of pain based on the type of urethral sampling material used. The plastic loop had the lowest pain scores, while the dacron gonorrhoea swab increased the pain scores by 14.5 mm more than the plastic loop and the rayon chlamydia swab by 28.7 mm more than the plastic loop. The raw pain scores with the rayon chlamydia swabs, especially when used as a second or third swab, were severe and comparable to severe pain experienced in other settings.7 8
Genitourinary symptoms such as dysuria and urethral discharge were not associated with higher pain scores nor was the finding of polymorphonuclear cells on the urethral slide. This was an unexpected finding as it was thought that symptomatic patients would have higher pain scores, but this study shows that having some background genitourinary pain did not translate into more pain from a urethral swab. This could be because these participants were already used to a certain amount of background discomfort from symptoms. Having a urethral swab performed in the past was associated with a reduction in the pain scores.
The study sample from the demographic data represents men presenting to the clinic for STI services apart from an excess of heterosexual individuals, so the results of the study should be generalisable (local data, not presented). The swab types used are also representative of the types of swabs available as they are both made of synthetic fibres and are 3 mm in size.
A limitation of the study is that the primary outcome is subjective as it is based on patients' interpretation of pain. The 100 mm VAS is, however, widely used in clinical trials. Another limitation of this study is the lack of blinding. It was not possible to blind the investigators or the patients as the three urethral sampling methods could not be constructed in such a way so as to look identical.
Nucleic acid amplification tests (NAAT) are the current standard of care for the diagnosis of genital chlamydial infections and can be performed painlessly on first void urine samples in men, with results that are just as good for the detection of chlamydia as urethral swabs.9 10 Urine tests either by means of NAAT, Gram-stained preparation from a centrifuged first-pass urine specimen or leucocyte esterase testing avoid the need for HCW-associated pain, which although generally of a short duration can be severe and may discourage the uptake of STI testing in the general population.11 12 Current national guidelines recommend that direct urethral sampling can be undertaken to diagnose urethritis and for the purposes of obtaining a specimen for Neisseria gonorrhoeae culture for antimicrobial sensitivity testing.12 13 If urethral swabs are to be taken then this study provides further support for keeping the number of directly obtained specimens from the urethra to the absolute minimum.
Direct urethral sampling is associated with discomfort and/or pain in men, which was significantly greater with a swab than with a plastic loop. Urine should therefore be the specimen type of choice. When direct urethral sampling is indicated, however, a loop is to be preferred to a urethral swab.
Key messages
Direct urethral sampling was associated with a median pain score on a 100 mm pain scale of 60.5 mm using a rayon chlamydia swab and 52 mm using a dacron gonorrhoea swab.
The rayon chlamydia swab increases the pain scores by 28.7 mm and the dacron gonorrhoea swab by 14.5 mm more than the plastic loop on a 100 mm pain scale.
Urine tests using NAAT, Gram-stained preparations from a centrifuged first-pass urine specimen or leucocyte esterase testing avoid the need for painful swabs.
Urine tests should be the specimen of choice; however, when direct urethral sampling is indicated a loop is to be preferred to a urethral swab.
References
Supplementary materials
Web Only Data sti.2010.042861
Files in this Data Supplement:
Footnotes
Funding The authors were funded as part of their clinical work.
Competing interests None to declare.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the Derbyshire research ethics committee, ref: 07/H0401/158.
Provenance and peer review Not commissioned; externally peer reviewed.