Evaluation of the Quick Wee method of inducing faster clean catch urine collection in pre-continent infants: a randomized controlled trial

Urinalysis and urine culture are both integral to the assessment of infants who are febrile or present with acute undifferentiated symptom complexes. The American Academy of Pediatrics (AAP) recommends invasive procedures, such as urethral catheterisation or suprapubic aspiration, to obtain urine samples in precontinent infants [1]. A study by Selekman based on the experience of 2726 parents of infants who underwent urinary catheterization reported that 56% of parents were extremely distressed by the procedure [2].

The NICE guidelines, which were updated in 2018, recommend a clean catch strategy for urine collection in pre-continent infants and children [3]. The time taken to obtain a clean catch urine sample can be considerable. In a study by Tosif of children aged 2–48 months, 139 (64%) had a successful clean catch urine specimen obtained; however, the median time taken was 30.5 minutes [interquartile range (IQR): 11–61 minutes] [4]. The guidelines also actively discourage the use of bag specimens to obtain urine samples.

Recently, the Quick Wee method of bladder stimulation, which uses a non-invasive strategy to induce voiding within 5 minutes, was evaluated by Kaufman in a tertiary care centre. In this evaluation, the Quick Wee method was an effective method for obtaining clean catch urine specimens in infants 1–12 months of age with the primary outcome being achieved in 31% of the intervention group versus 12% of the control group [P < 0.0001, absolute difference 19%, number needed to treat = 5 (NNT)] [5].

Prior to undertaking the present study, we utilized urine bags in precontinent infants to obtain urine samples because it is customary and a matter of practice not to use invasive procedures routinely to obtain urine in Ireland. Bag specimen urine collection had been well established in our unit and was seen as both convenient and time-efficient by the nursing staff. The Quick Wee method, if effective would allow our practice to be modified in keeping with current NICE guidelines. We therefore undertook a replication study of the Quick Wee method of urine collection in our secondary care institution to evaluate the effectiveness of this method.

Ethical approval was obtained from the Research Ethics Committee of Mayo University Hospital, Ireland prior to initiation of the trial (REC approval number 20181001). The trial was registered with the ICTRN registry (ISRCTN43796385).

We performed a prospective, randomized, non-blinded replication trial between February and July 2019 in both the Emergency Department and Pediatric Decision Unit of Mayo University Hospital, which provides secondary pediatric care with 7500 emergency paediatric presentations per year.

Patient flow through the study is detailed in Fig. 1. The parents of 153 infants were approached to take part in the study, and 142 parents agreed. 140 infants were included in an intention-to-treat analysis, with 73 in the intervention group and 67 in the control group. One infant in the intervention group was excluded prior to analysis because the age criterion was breeched. There was one withdrawal post-randomization prior to starting the trial period in the control group. Two infants were randomized to the control group but had the intervention performed due to human error. These were included in the intention-to-treat analysis.

The baseline characteristics were similar in both groups. These are detailed in Table 1. Comorbidities of included infants were recorded by enrolling nurse/doctor. Intervention group: laryngomalacia (n = 1), mild hydronephrosis (n = 1), hydrocephalus with ventriculoperitoneal shunt (n = 2); Control group: prematurity (n = 1), hydrocephalus with ventriculoperitoneal shunt (n = 2), vaccinations in prior 24 h (n = 1), developmental dysplasia of hip (n = 1), trisomy 21 (n = 1). The clinical indications for urine collection are outlined in Table 2. Clinicians could list more than one indication.

In the study, recruitment was performed by 113 (81%) doctors and 27 (19%) nursing staff. Standard cleaning was performed by 91 (65%) doctors, 38 (27%) nursing staff, and 11 (8%) parents. Catching of urine samples in a sterile container was performed by 44 (32%) doctors, 31 (22%) nursing staff, and 65 (46%) parents. The intervention, rubbing of cold saline to the suprapubic area, was performed by 49 (67%) doctors, 19 (26%) nursing staff, and 5 (7%) parents.

Urine samples containing mixed bacterial growth on culture were equal between control and intervention group, 54% (6 of 11) in the intervention group compared with 60% (3 of 5) in the control group. Overall, nine (31%) of urines collected in the study were contaminated.

The study was accepted well by both parents and clinical staff. There were no significant difficulties reported in either group. On a five-point Likert scale (very unsatisfied to very satisfied), the mean score for parents was 4.16 with a positive skew (percentage of respondents choosing point 4 or 5 of scale) of 81% and mean score for clinicians was 4.35 with a positive skew of 83.6% (Fig. 2).

The present study was a replication study of the Quick Wee method of urine collection in pre-continent infants from 1 month to 1 year of age and was carried out in a hospital providing secondary care [5]. We did not observe a statistically significant difference in urine collection obtained using this method. The success rate was 18% in the control group and 25% in the Quick Wee group (Absolute difference 7%, 95% CI − 7% to 20%, NNT 14). In the original study by Kauffman, the success rate was 12% in the control group and 31% in the Quick Wee group.

To date and to our knowledge, there have not been other studies registered to evaluate the Quick Wee method against standard methods of urine collection. It is possible that non-registered trials have been performed that did not show statistically significant results. Such studies are less likely to be published [7‐9].

Failure to replicate the results of original trials has been shown across many areas of medicine. Ioannidis et al. showed that 44% of highly cited studies had results inconsistent with a replication attempt [10]. The author cites differences in the disease spectrum, population, eligibility criteria, and the use of concomitant interventions as potential explanations.

The Quick Wee method is based on the physiology of micturition in young infants. Stimulation of the voiding reflex can be utilized to obtain urine samples prior to the development of the control of voiding, which normally develops at approximately 2 years of age via the pudendal nerve [11]. This physiological feature can be provoked by stimulating the bladder or perineal skin [11, 12]. Cold stimulation will trigger parasympathetic detrusor contraction through the somato-bladder reflex mechanism and can similarly be utilized [12‐14]. This reflex may be temperature sensitive. The temperature in Castlebar during the 5-month study period ranged from 5 to 15 °C compared to Melbourne which ranged from 15 to 21 °C during a 10-month study period. The level of dehydration or illness in the cohorts in our study was similar to those in the study by Kaufman.

Parents could undertake specific tasks in this study. 8% (11) of parents undertook standard cleaning, 7% (5) performed the suprapubic rubbing with cold saline, with two infants voiding, and 46% (65) were assigned to urine collection. This active engagement may have contributed to the high parent satisfaction rating, and it reflects current pediatric practice of partnership with parents in the care of their children. Parental involvement, especially in urine collection, is essential in clinical practice.

In our study, the impact of specific potential influencers included: (1) the interval from the infant’s last void; (2) the impact of feeding directly before the collection of urine, and (3) the infant’s age were not evaluated. The frequency of voiding in infants is quoted in the literature as four times per 4-h period up to 6 months of age and three times per 4-h period for ages 6–12 months [11, 14]. It is plausible that a longer time period since the infant’s last urine void and the administration of fluid (e.g., 20 mL/kg) would be expected to result in more urine being present in the bladder. This could be assessed utilizing point-of-care ultrasonography. This should then theoretically result in reflex emptying on application of the Quick Wee method. These factors need to be assessed in future studies.

A post hoc subgroup analysis explored differences in utility of the method based on age, given the changing physiology of infant micturition over the first year of life. We saw a larger positive response to this method in the group over 6 months (absolute difference of 4% with NNT of 23 in the 1–6-month group compared with a 16% absolute difference with number needed to treat of 7 in the over 6-month group). These results must be interpreted with significant caution because we did not power the study sufficiently for this outcome and because the absence of stratification based on age prior to randomization led to a significant difference in group size. In our study and in Kaufmann’s, the percentage of patients had not voided at 5 minutes were similar (75% and 69%)[5].

Our current practice is to actively encourage waiting, until the infant voids spontaneously, because it is customary and common practice not to routinely perform infantile catheterization. We actively dissuade the use of bag specimens for urine collection. Other non-invasive strategies can be used to obtain urine samples from pre-continent infants, as reported by Herreros Fernandez, et al [15]; however, the strategy relies on a combination of specific fluid intake and non-invasive bladder stimulation [15]. This process consists of the following steps: (1) the administration of 25 mL/kg of fluid, if the infant is older than 2 weeks; (2) 25 minutes post feeding, the genitals are cleaned thoroughly with warm water and soap and dried with sterile gauze; (3) prior to commencing the non-invasive bladder stimulation the infant receives non-pharmacological analgesia, such as 2% sucrose or non-nutritive sucking to prevent or reduce crying, and (4) the infant is held under the armpit with the legs dangling and tapping of the suprapubic area at a frequency of 100 taps per minute for 30 seconds, followed by lumbar stimulation manoeuvres (light circular massage in the lumbar paravertebral region for 30 seconds) commences. This process is continued for 5 minutes, or until the infant voids. This method requires specific training to ensure competence and requires non-pharmacological analgesia to reduce infant distress levels. The original cohort study involved 80 infants under 1 month of age, and 69 (86.3%) infants voided within 5 minutes.

This study was replicated by Crombie in a cohort of 147 infants < 90 days old [16]. The success rate of induced voiding was lower at 53.1%. Both studies utilized fluid administration and bladder stimulation but the relative contributions of these interventions were not assessed. Another study by Tran modified the bladder and lumbar stimulation protocol in non-ambulant infants. Children < 2 years were recruited, and the stimulation time was reduced to 3 min [17]. If voiding did not occur, infants had oral fluids administered prior to a second attempt. Of the inception cohort of 142 patients, 60 (42.2%) voided within 3 min. Of the 82 who received fluids and had a second episode of stimulation, 19 (23.1%) infants voided. A key finding of this study was the impact of age on the primary outcome. Of the 27 infants < 1 month old, 24 (88.9%) voided as opposed to 2 (28.6%) of the 7 participants > 1 year of age. Cohort studies can overestimate effect sizes and consequently, an RCT by Demoncy to define the impact of bladder stimulation, as a non-invasive technique to collect urine, to diagnose UTI in infants < 6 months is currently underway [18].

The main strength of our study was adherence to the core elements of an RCT with true randomization and allocation concealment. Our study was adequately powered to detect the 20% clinical difference that we sought to modify our clinical practice. We used a simple technique, which is not resource-intensive when compared to the bladder and lumbar stimulation technique described above and used in other studies, to address a common pediatric challenge. The high levels of acceptability of this method will aid practice modification.

Limitations of this study include: (1) not stratifying infants by age, i.e., 1–6 months and > 6 months; and (2) not determining the time of the last void by infants prior to study enrolment.

In conclusion, this prospective, randomized, non-blinded superiority trial did not show a statistically significant benefit for carrying out the Quick Wee method of inducing urine collection in pre-continent infants.