1. Introduction
Urinary tract infection (UTI) is one of the most common bacterial infections among infants, with a prevalence of 5% - 7% in children below age 24 months with a fever of unknown source [1] [2] [3] [4] [5]. Infants are at higher risk in the first few months of life due to their immune system, that is, not fully developed [6]. The prevalence and incidence of UTI are higher among boys during the first three months and higher among girls during the first year of life [5] [6]. In boys without circumcision, an increase in the frequency of UTI is noted, with a risk of developing UTI is 10 to 12 times higher during the first six months of life [6]. Circumcision has shown a protective effect by reducing the odds of UTI by 85% [7]. It is essential to identify febrile infants with UTI as they often present with non-specific symptoms, typically an unexplained high fever [1] [8]. An accurate diagnosis of UTI requires a good quality urine sample for urine culture and other screening tools. With an early accurate diagnosis, treatment can be initiated as soon as possible, meanwhile avoiding unnecessary therapies [3] [9] [10]. Early diagnosis is crucial to prevent progression to chronic complications such as pyelonephritis leading to renal scarring, renal failure, and hypertension [1] [5] [9] [11] [12].
Urine sample collection methods in infants are commonly classified as non-invasive and invasive. Examples of non-invasive methods are urine bag, clean-catch urine (CCU) and urine collection pads; and the invasive methods, namely catheterization and suprapubic aspiration (SPA). These methods are quoted in the Paediatrics Protocol for Malaysian Hospitals, Clinical Practice Guidelines of the Royal Children’s Hospital (RCH) in Australia, National Institute for Health and Care Excellence (NICE), and American Academy of Paediatrics (AAP) [13] [14] [15] [16]. In all four guidelines, urine bag collection is the least recommended due to its high contamination rate. Paediatrics Protocol, RCH and NICE recommend CCU as the first-line method [13] [14] [15]. Only NICE considered urine collection pads in their guideline as to the second alternative for a non-invasive method before proceeding to make an attempt with invasive methods [15]. Invasive methods are more preferred for obtaining a clean and reliable sample [6]. However, these techniques cause pain, have higher the risk of complications, and may cause distress in parents [17]. Catheterization has a sensitivity of 95% and specificity of 99% as compared to SPA [13] [16], and the contamination rate as low as 10% [14]. SPA is known as the “gold standard” in both Paediatrics Protocol and the RCH guidelines [13] [14]. It yields the most reliable sample and lowest contamination rate of 1%, at the same time, it is the most technically challenging method [6] [14].
Mamta et al. (2019) stated that invasive methods have a high failure rate in newborns due to their anatomical characteristics and irregular voiding pattern [18]. In the AAP guideline, the urine sample collection method is determined clinically. If a febrile infant is in a very ill condition, highly suspected for UTI and requires immediate antimicrobial therapy, urine samples must be collected via catheterization or SPA. Urine culture ordered before administration of antimicrobial therapy is crucial so that a definitive diagnosis of UTI can be established, and this is to prevent a missed diagnosis leading to long term sequelae if untreated, or an overdiagnosis causing overtreatment in infants. Urine samples collected via urine bag will not be able to confirm a diagnosis of UTI due to its high false-positive rate of 88%, which is not reliable [16]. A midstream CCU sample is more challenging to obtain from infants below the age of two as compared to older children who are toilet trained and can void voluntarily. Children are developmentally ready for toilet training between 18 months to 2.5 years old [19].
Fernandez-Herreros et al. (2013) proposed a new technique that incorporates bladder stimulation manoeuvres to facilitate the collection of CCU samples. Infants will receive a bladder stimulation via gentle tapping at the suprapubic area, followed by stimulation of the lumbar paravertebral zone with a light circular massage given at the lower back. The study showed a high success rate, fast sample collection, and no complications occurred with this new technique [20]. Many studies have been carried out on bladder stimulation techniques [1] [2] [4] [8] [9] [10] [11] [17] [18] [20] [21] [22]. Based on these available studies, bladder stimulation techniques manifested a good potential as a practical alternative to the two widely practiced methods in the current clinical setting in Malaysia, i.e. standard CCU that is time-consuming and catheterization that is relatively more invasive to infants. Nevertheless, insufficient data is available to evaluate its feasibility in a Malaysia clinical setting as no relevant studies were conducted yet in South East Asia, including Malaysia itself.
That is why this literature review has been conducted aiming to compare various non-invasive urine sample collection methods which could be more feasible for a district hospital setting in Malaysia. The non-invasive methods to be studied are standard CCU (without any additional stimulation), bladder stimulation techniques, urine collection pads, and urine bag.
2. Methods
The two databases, Google Scholar (n = 257) and PubMed (n = 206), were used in searching and identifying potential studies published from 2002 to 2020. The Medical Subject Headings (MeSH) terms “infants” AND “urinary tract infection” AND “urine specimen collection” were used to search relevant studies, generating 463 search results. Duplicates articles were removed by using Endnote citation manager software and narrowing the search results to 114 studies. Titles and abstracts of articles in the initial search results were screened according to inclusion and exclusion criteria, leaving 31 studies eligible. Following that, full-text articles of eligible studies were retrieved and further assessed. Finally, 17 studies (as shown in Figure 1 and Table 1) were included and data were extracted based on three major variables (success rate, sample collection time and contamination rate) for the analysis in this literature review.
![]()
Figure 1. PRISMA chart used for the selection of articles.
Studies involved children aged below 24 months or 2 years, compared to any non-invasive urine sample collection methods and published in English were included in this literature review. There was no restriction for study design; any study reported one of the following outcomes was included: success rate of urine sample collection, time taken for urine sample collection, and contamination rate. Exclusion criteria were adult age group, non-English reports, only comparing invasive urine sample collection methods, and none of the above outcomes measured.
While extracting data, MPLC and NKJ1 worked independently as the main reviewers. They extracted data in a standardized form (Table 1) which included the following information: study information like list of authors, country and year of publication, study design and sample size, main study findings and conclusion. This form was specifically developed for this review purpose in order to minimize human errors and biases. The third reviewer, either CGY or NKP2, contributed when the main reviewers could not come to a general agreement. The whole review process was monitored and supervised by INCbS and YCY3 to ensure the quality.
3. Results
We included 17 studies in this review, and extracted data were grouped based on the outcomes (success rate, sample collection time and contamination rate) measured. A brief description of the techniques for each non-invasive urine sample collection method, compared in this review, is outlined below. Prior to each episode of collection of urine regardless of which non-invasive method, the genital area of infants is cleaned with soap and or water [4] [20] [23].
3.1. Standard CCU
A sterile urine container is prepared and ready to collect urine from the infant [23]. Without any additional stimulation or manoeuvres, the clinician, parent, or carer will wait until the infant to void spontaneously [4]. Urine container is not supposed to be in contact with the perineum [23].
3.2. Bladder Stimulation Techniques
This method collects a CCU sample with the same steps as a standard CCU but complemented with bladder stimulation techniques. Among the 17 studies included, two types of bladder stimulation techniques were described: Quick-Wee method and lumbar/bladder stimulation method. In the Quick-Wee method, a rub is given at the suprapubic area of an infant in a circular motion with gauze soaked in cold saline. A disposable plastic forceps is used to hold the gauze to maintain sterility of procedure [4]. The suprapubic rub is continued until a CCU sample is obtained or the duration of 5 minutes is reached [1] [4] [21].
The lumbar/bladder stimulation method proposed by Fernandez-Herreros et al. (2013) requires infants to be fed either breastfeeding or formula intake 25 minutes before urine collection commences. The genital area is cleaned, and non-pharmacological analgesia is given to infants prepared for urine collection. This stimulation technique needs at least two trained personnel, one holding the infant under the armpits with legs dangling, while another performs bladder stimulation followed by lumbar paravertebral zone stimulation. Bladder stimulation is done by gently tapping the suprapubic area at a rate of 100 taps per minute over 30 seconds. Then, a lumbar paravertebral zone stimulation is conducted by applying light circular massage at the lower back of the infant for 30 seconds. The two manoeuvres are repeated until a CCU sample is obtained [20].
3.3. Urine Collection Pads
The commonly used pads are Newcastle Sterile Urine Collection Packs (alternatively known as Euron Uricol Urine Collection Packs) [23] [24] [25] [26] [27]. Each pack consists of an instruction leaflet, two soft non-woven sample urine collection pads, a syringe and a specimen bottle. It is a urine sample collection system named after Newcastle University, which developed and supported by its clinical research [26] [27]. The urine collection pad is placed inside the infant’s nappy. Urine is retrieved by aspiration using a syringe. To detect if the infant has voided, pad is checked every 10 or 30 minutes, or enuresis alarm is utilized. Pads soiled with stool will be discarded and replaced with a new pad to repeat urine collection [23] [24] [25]. In a study conducted by Alam et al. (2005), Newcastle pads were also compared with commercial sanitary pads [23].
3.4. Urine Bag
Urine bag is attached at perineum of an infant by a trained nurse using the standard perineal cleansing procedure [3] [23] [28]. The urine bag is left until urine is collected or approximately for one hour without replacement unless indicated by leakage, stool contamination or detached from the skin [3]. After peeling off the urine bag gently, a collected urine sample will be poured into a sterile urine container and sent to the laboratory within 20 minutes to 1 hour [3] [23] [28].
3.5. Success Rate of Urine Sample Collection
A successful urine sample collection is defined as urine collection within 5 minutes [1] [2] [8] [10] [11] [17] [18] [20] [22]. Out of seven, six studies revealed a higher overall success rate (ranging from 49% to 88.9% with a statistically significant difference) of urine sample collection in lumbar/bladder stimulation method as compared to standard CCU methods [1] [2] [8] [11] [18] [20] [22]. Tran et al. (2016) modified the study adapted from Fernandez-Herreros et al. (2013), where the fluid is introduced after the failure of the first attempt and before the second attempt with the duration apart of 30 minutes (instead of prior to commencement of procedure). A sterile bag is placed onto the child in case of the infant passes urine before the second attempt. The success rate in the first attempt and second attempts was 42.3% and 23.3%, respectively [8].
Valleix-Leclerc et al. (2016) however reported an overall success rate is as low as 27% [10]. Kapoor & Mekle (2017) presented a success rate of 88.3% in the lumbar/bladder stimulation method, which is higher than its control group using a sterile adhesive bag (success rate 86.67%) [17]. Kaufman et al. (2017) and Hall-Million & Howard (2017) studied on the Quick-Wee method and found significantly higher success rates in contrast to standard CCU [1] [4] [21]. The success rate of urine sample collection has a very weak association with gender. Five studies with lumbar/bladder stimulation as an intervention found no statistically significant difference between boys and girls in terms of the success rate [1] [2] [3] [8] [10] [22]. Apart from the systematic review conducted by Chandy et al. (2020) that reported no significant association, three studies had shown that the younger age group results in a significantly higher success rate [1] [2] [8] [10].
In the study conducted by Labrosse et al. (2016), a higher proportion of success was seen in infants below 89 days old than those aged above 89 days [2]. Tran et al. (2016) presented a success rate of 88.9% in newborns and 28.6% in infants older than one year (p = 0.0001) [8]. A statistically significant decrease in success rate was noted with the weight of infants [8] [10]. The prospective non-controlled study conducted by Valleix-Leclerc et al. (2016) included 48 infants with a median weight of 8.7 kg. In this study, despite the low overall success rate of 27%, among children aged below three months with a median weight of 5.2 kg, the success rate of urine sample collection was 50%. Weight was a significant factor resulting in difficulties in positioning an infant [10].
There was no significant association of success rate with low oral intake and voiding within the hour before urine collection [2]. Discomfort during urine collection with the lumbar/bladder stimulation method was reported at least once throughout the urine collection in 58.5% infants (95% CI = 50.4 - 66.6), with a significantly increased prevalence with age (p = 0.01) and weight (p = 0.012) [8].
3.6. Time Taken for Urine Sample Collection
Eight studies of lumbar/bladder stimulation that measured the time taken for urine sample collection showed significantly faster collection than the control groups (standard CCU, adhesive urine bag, or invasive methods). The mean time for urine sample collection ranged from 57 seconds to 2 minutes, and the median time ranged from 45 seconds to 60 seconds among the intervention group using the lumbar/bladder stimulation method [2] [8] [10] [11] [17] [18] [20] [22]. In the study by Altuntas et al. (2015), the median time for urine sample collection in the control group using standard CCU was 300 seconds (p < 0.001) [22].
Kapoor & Mekle (2017) presented a median time of 70 minutes for urine sample collection via adhesive urine bag, which is significantly longer (p < 0.001) than the lumbar/bladder stimulation where the median time was 55 seconds [17]. For urine collection pads, the minimum median time for urine sample collection was 45 minutes [24] [25]. Rao et al. (2004) compared single use of urine collection pads versus urine collection pads replaced every 30 minutes until a sample is obtained, revealed the median time of 80 minutes and 45 minutes, respectively, but no significant difference was found between these two groups (p = 0.056) [24]. Farrell et al. (2002) conducted a pilot, method comparison study to assess the reliability of urine collection pads in the microbiological examination. Urine samples were collected with urine bags followed by urine collection pads from all infants in the study population. A median time of 62.5 minutes was recorded for the entire urine collection process [25].
Studies on the Quick-Wee method did not record the time taken for urine sample collection; however, urine samples were collected within 5 minutes in 30% of the infants [4] [21]. As such, it can be said that among the four non-invasive methods, bladder stimulation technique was the fastest method, followed by standard CCU, and similar time is taken for urine collection pads and urine bags.
The association between gender and time taken for urine sample collection was evaluated in three studies. Fernandez-Herreros et al. (2013) reported a shorter mean and median time of urine sample collection in females than males using the lumbar/bladder stimulation method [20]. Kumar & Nithin (2019) reported the opposite result of a shorter mean time in males than females [11]. Nonetheless, both studies found no statistically significant difference between males and females in the time taken for urine sample collection [11] [20]. Mamta et al. (2019) compared the median time taken of urine sample collection between males and females in both methods, i.e. of lumbar/bladder stimulation (intervention group) and standard CCU (control group). They reported the shortest median time in males of the intervention group, followed by females of the control group, males of the control group and females of the intervention group. In the intervention group, a significant difference was noted between males and females in the time taken for urine sample collection (p = 0.008) [18]. Kapoor & Mekle (2017) studied on the duration of the last feeding and found no correlation with the time taken for urine sample collection (p = 0.57) [17].
3.7. Contamination Rate
Six studies on bladder stimulation techniques measured the contamination rate as one of the outcomes [1] [2] [4] [18] [21] [22]. Mamta et al. (2019) and Altuntas et al. (2015) compared the lumbar/bladder stimulation with the control group using standard CCU method [18] [22]. An overall contamination rate in the whole population of 1.9% was reported by Mamta et al. (2019), which occurred in one of the infants in the control group who failed to give a urine sample [18]. Altuntas et al. (2015) reported a contamination rate of 24% and 29% in the intervention group and the control group, respectively (p = 0.77) [22]. A prospective cohort study by Labrosse et al., (2016) which was also included in a systematic review by Chandy et al. (2020), revealed no statistically significant difference in the contamination rates between lumbar/bladder stimulation versus invasive methods of 16% and 6%, respectively [1] [2].
Another cross-sectional study included by Chandy et al. (2020) that compared lumbar/bladder stimulation and invasive methods too reported a low contamination rate in both groups, in addition, high sensitivity of 97% and specificity of 89% [1]. Studies of the Quick-Wee method showed low contamination rates in contrast to standard CCU and urine bag [1] [4] [21]. The contamination rate in the Quick-Wee method was 27% versus 46% in standard CCU but with no statistically significant difference (p = 0.29) [4] [21].
Chandy et al. (2020) however presented a statistically significantly lower contamination rate in the Quick-Wee method of 7.7% than 51% in urine bag from a non-randomized trial included in the systematic review [1]. Alam et al. (2005) conducted a study on standard CCU, urine collection pads, and urine bags, and found a significantly lower contamination rate in standard CCU (14.7%) than both urine collection pads (29%) and urine bags (26.6%). There was no statistically significant difference between urine collection pads and urine bags. This study also compared two types of urine collection pads: commercial sanitary pads and Newcastle pads. The contamination rate was found to be lower in Newcastle pads (26%) than commercial sanitary pads (36%), but there was no statistically significant difference [23]. Li et al. (2002) screened 100 infants who were followed up for previous UTI. Urine samples were collected via urine bag for urine culture, resulted in 95 of them with positive culture due to contamination and the remaining 5 with genuine bacteriuria. They reported a false-positive rate of 36.8%, a specificity of 63.2%, and positive predictive value of 12.5% [28].
Another study by Karacan et al. (2010) assessed urine bag and standard CCU with invasive methods and found the highest contamination rate of 43.9% in urine bag (p < 0.001), followed by standard CCU and catheterization (both 14.3%), and 9.1% in SPA [3]. Herreros et al. (2018) conducted a study evaluating the accuracy of the urine dipstick test in CCU samples as a UTI screening modality. There were 5% of CCU samples and 8% of samples obtained by catheterization contaminated, therefore excluded from the study [9].
In short, there was no statistically significant difference between bladder stimulation techniques and standard CCU in contamination rate, but a standard CCU provided a significantly lower contamination rate than urine collection pads and urine bags. There was also no significant correlation between gender and contamination rates of a urine sample [22] [23]. In the study by Altuntas et al. (2015), out of the entire study population (n = 127), the contamination rate in females was 34.3% and 20% in males (p = 0.86) [22]. Alam et al. (2005) found no statistically significant difference between males (12%) and females (19%) of infants using standard CCU method in the contamination rate (p = 0.3) [23].
3.8. Complications and Adverse Events
Apart from consolable crying, no other complications or adverse events were reported in almost all non-invasive urine sample collection methods [1] [10] [17] [18] [20] [22]. Local skin redness at sites of manoeuvres applied and discomfort in infants were noted in the lumbar/bladder stimulation method [1] [8] [10].
3.9. Satisfaction of Clinicians, Parents and Preference of Patient
Only three studies that adopted the Quick-Wee method evaluated the satisfaction of clinicians and parents. The Quick-Wee method is statistically significantly more favoured as compared to the control group using the standard CCU method [1] [4] [21]. Karacan et al. (2010) studied the preference of patients and found that a sterile urine bag was the most preferred method among children in all three age groups of newborns, 1 to 6 months, and 7 to 24 months old. Other methods compared in this study were standard CCU, catherization, and SPA [3].
4. Discussion
Based on this extensive review, bladder stimulation techniques are the ideal method due to its high success rate, fast collection and low contamination rate among the four non-invasive urine sample collection methods. Standard CCU is indeed a useful method of urine sample collection for UTI diagnosis. It has a contamination rate as low as catherization. However, it is less time-efficient where the duration of urine sample collection varying in each infant may be unpredictable. Urine collection pads and urine bags not only are time-consuming but high in contamination rate, further reduced its favourability, especially if indicated for urine culture [3] [11] [20] [21] [23] [28].
The lumbar/bladder stimulation method is found to have a significantly high success rate and fast collection of urine samples in infants. Unlike older children, newborns are yet to develop voluntary inhibition of spinal micturition reflex that control the contraction of detrusor muscles in the urinary bladder via the cerebral cortex. Fernandez-Hererros et al. (2013) proposed that this reflex can, however, be triggered in newborns [18] [20]. The success rate of urine sample collection was higher in younger age groups [4]. Spinal micturition reflex influenced by cortical feedback matures over time. By the age of three, voluntary control of periurethral striated muscle sphincter will usually be gained in the child [4] [10] [17].
Weight was another major limitation as presented by Valleix-Leclerc et al. (2016), contributing to difficulties in positioning the infant appropriately. The study may have reported no significant difference in terms of age, but considering the strong association between age and weight, this may justify a lack of power in the study instead [10]. Other limitations addressed in the studies were the absence of a control group, pre-requisite of two to three trained staffs which may be less feasible in heavier infants, fluid intake not standardized, an infant with low oral intake not excluded, and precautions to maintain sterility was not done [2] [4] [20].
The Quick-Wee method had a significantly higher success rate of collecting a urine sample within 5 minutes as well. It is easy to be carried out without any special equipment needed, hence can be adopted in clinical settings with low resources [4]. Nevertheless, Hall-Million & Howard (2017) doubted if a similar study designed as such could be delivered in a setting where urine sample collection is routinely done by nurses since no reports on reliability among the team in present studies are available [21]. The studies only included infants of 1 to 12 months of age but not neonates and non-toilet-trained children above 1-year-old. Therefore the findings may not be applicable across all paediatric age groups [4] [21]. Kaufman et al. (2017) addressed the issue of lacking a microbiological definition of contamination as one of the limitations in the study [4]. The difference in contamination rate that was not significant may be due to a small number of culture results available [21].
5. Conclusions
In summary, bladder stimulation techniques are the most favoured among all non-invasive urine sample collection methods. Limitations in the studies, as addressed above, require further research to ensure the reliability of urine sample collected so it could be turned into a common practice in the clinical setting. Further studies on lumbar/bladder stimulation as proposed by Fernandez-Herreros et al. (2013), the Quick-Wee method, as well as invasive methods, are also worth to be considered, measuring the same outcomes: a success rate of urine sample collection, time take for urine sample collection and contamination rates. From this literature review, the bladder stimulation technique is perceived as more feasible for practice in a district hospital setting in Malaysia. There is a higher likelihood to implement this stimulation technique in a tertiary setting if research on new technique is found feasible in a less resourceful district hospital setting.
Currently, many studies have compared bladder stimulation techniques with other non-invasive or invasive methods, but not between the two bladder stimulation techniques. Invasive methods are suggested as a control group since catheterization is more widely practiced than standard CCU for its practicality and time-efficiency in busy hospital settings in Malaysia, despite the fact that standard CCU is the first choice recommended in the Paediatrics Protocol for Malaysia Hospitals [13].
List of Abbreviation (Sorted Alphabetically)
AAP American Academy of Paediatrics
CCU Clean-catch urine
MeSH Medical Subject Headings
NICE National Institute for Health and Care Excellence
RCH Royal Children’s Hospital
SPA Suprapubic aspiration
UTI Urinary tract infection
NOTES
1MPLC: Michelle Pei Ling Chia; NKJ: Nowrozy Kamar Jahan.
2CGY: Christina Gertrude Yap; NKP: Naganathan Kathiresan Pillai.
3INCbS: Intan Nor Chahaya binti Shukor2; YCY: Yee Chau Yen.