Advice

Clinical and technical evidence

Clinical and technical evidence

A literature search was carried out for this briefing in accordance with the interim process and methods statement for medtech innovation briefings. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.

Published evidence

The evidence from 1 randomised comparative study of 600 women in the UK, 2 cohort studies of 246 people in Belgium, and 2 cross-sectional studies of 113 people in Spain and 25 people in Belgium were included in the briefing.

Three abstracts were identified in the search. They were not included because:

The clinical evidence and its strengths and limitations is summarised in the overall assessment of the evidence.

Overall assessment of the evidence

One UK single-centre comparative study (Cadman et al. 2021) compared a urine sample using Colli‑Pee and 4 vaginal sampling devices, including wet and dry transport methods for HPV detection in 600 women who had a positive result after cervical screening. The study reported no differences in detection rates of high‑risk HPV and sensitivity for detecting cervical intraepithelial neoplasia grade 2+ and cervical intraepithelial neoplasia grade 3+.

One single-centre study (Baetselier et al. 2019) included people at high risk of HIV and compared the molecular detection of sexually transmitted infections using home-collected urine samples and urine samples collected in a clinic. The agreements between 2 types of samples were over 95% for detecting chlamydia, gonorrhoea and Mycoplasma genitalium. The study also reported their experience using the Colli‑Pee device.

Pattyn et al. (2019) included a small number of people who self-reported and had tested positive for HPV based on cervical samples (n=31). A total of 12 were found HPV positive based on their urine sample. The study analysis suggested that there was no significant effect of the timing of collection (morning compared with later during the day) on copies of HPV DNA detected. Colli‑Pee collected urine samples show higher HPV concentrations than cup‑collected samples. However, at high concentrations of HPV DNA, the benefit of Colli‑Pee disappears.

Of 2 cross-sectional studies, Leeman et al. (2017) compared the sensitivity of detecting cervical intraepithelial neoplasia grade 2+ in self-collected urine samples, self-collected cervicovaginal samples and clinician-taken smears. The results were comparable. Téblick et al. (2021) compared the detection of human and viral outcomes using different volumes for the first urine samples, suggesting limited effects of collection volumes on human and HPV DNA endpoints.

Overall, there is limited published evidence on Colli‑Pee. The evidence available is helpful to see the potential of the technology, particularly the sensitivity for detecting HPV reported in an abstract using samples by different methods. However, a well-done diagnostic study would be helpful to gain more certainty in terms of the diagnostic accuracy using Colli‑Pee against standard reference methods used in the NHS and potential savings.

Cadman et al. (2021)

Intervention and comparator

A stream urine sample was collected using the Colli‑Pee collection kit, which collects a 20 ml sample, of which 7 ml is a prefilled urine conservation medium (UCM).

Vaginal samples were collected either as a wet sample or a dry sample (dry flocked swab [DF], Qvintip [QT] and HerSwab [HS]).

Key outcomes

The rates of detecting high-risk HPV positivity were 76.0% and 71.7% for urine samples (n=504) and wet samples (n=300) respectively. There was no statistically significant difference.

Similar sensitivities for cervical intraepithelial neoplasia grade 2+ and cervical intraepithelial neoplasia grade 3+ were seen for samples collected by a dry swab, a wet swab or a urine sample.

  • The sensitivity for cervical intraepithelial neoplasia grade 3+ was 91.2% by wet sample, followed by 89.7% by a urine sample, 88.2% by DF, 81.8% by QT, and 77.4% by HS.

  • The sensitivity for cervical intraepithelial neoplasia grade 2+ was 90.0% by wet sample, followed by 88.6% by DF, 87.0% by a urine sample, 82.5% by QT and 82.0% by HS, with a specificity of 34.2%, 33.8%, 27.4%, 34.3% and 39.5% respectively.

Results of a survey evaluating women's preference for sample methods, reported that 46% women (n=208) preferred one of the 2 non-urine (vaginal) devices, 32% preferred urine (collected using Colli‑Pee; n=145), and 22% expressed no preference (n=101). Women found urine easiest to collect and were more confident they had taken the sample correctly compared with other vaginal samplings.

Strengths and limitations

This UK study included 600 women. It is a single-centre study. All women had urine samples. Women were randomised to collect vaginal samples as a wet sample or a dry sample. Time intervals between urine and vaginal samples were not reported in the study.

Téblick et al. (2021)

Intervention and comparator

Home-collected first void urine samples using Colli‑Pee with collector tubes that differ in volume (4 ml, 10 ml, 20 ml). Each collector tube was prefilled with UCM.

Key outcomes

There were no significant differences between the 3 first void urine collection volumes for detecting human endpoints including glyceraldehyde 3‑phosphate dehydrogenase, beta‑actin, and beta‑globin. More high‑risk HPV infections were detected using Colli‑Pee for the collection of 10 ml compared with 4 ml and 20 ml, but the differences were not statistically significant.

Strengths and limitations

The study had a small sample size. The author noted that all participants self-reported to have a high‑risk HPV infection in the previous 6 months, but this could not be verified by the study team.

Baetselier et al. (2019)

Intervention and comparator

People who agreed to take part in the study received a Colli‑Pee device and a prepaid envelope. They were instructed to collect first void urine the next day at home using the Colli‑Pee device (the home-based sample), to document the date and time of collection and to send the collector tube filled with urine back to the laboratory by regular post, using the prepaid envelope.

People also had urine samples collected during their visits to the clinic for testing for chlamydia, gonorrhoea, Mycoplasma genitalium and Trichomonas vaginalis.

Key outcomes

A total of 471 home-based samples from 213 people were received and included in the study. Trichomonas vaginalis was not detected. Percent agreement (Cohen's kappa coefficient) between home-based samples and clinical-based samples for chlamydia, gonorrhoea and Mycoplasma genitalium were 99.1% (0.75), 99.6% (0.87) and 98.3% (0.85) respectively.

A total of 164 people provided feedback about the use of the Colli‑Pee device. Nearly 88% of people found that the Colli‑Pee device was easy to use. Four people found the Colli‑Pee difficult to use (2.4%) and another 4 people found it difficult to follow the instructions (2.4%). People liked Colli‑Pee because it:

  • is easy to use (54.9%)

  • has no interruption of the urine flow (15.9%)

  • is hygienic (11.6%)

  • offers privacy of the home-based sample collection (11.0%).

People disliked the device because it:

  • is not recyclable (14.6%)

  • is not hygienic (10.4%)

  • is too large (6.1%).

Strengths and limitations

This is a single-centre study. Most of the home-based samples (79.6%) were taken within 2 days after the clinical-based samples, and 3.8% were taken after 20 days. Not all people who used a Colli‑Pee device completed the survey. The number of clinic and home-based samples received during the study decreased over time. The study was supported by Novosanis who provided the Colli‑Pee devices and partially paid for the analyses done on the home-based samples.

Pattyn et al. (2019)

Intervention and comparator

People who agreed to take part in the study received a package by postal mail containing 4 Colli‑Pee devices. They also received standard urine cups to collect first void urine samples. They were asked to sample the very first void urine of the day (U1) and a first void urine sample later during the day (U2) for 4 consecutive days.

The Colli‑Pee device captured a fixed volume of 13 ml (plus or minus 2 ml) first void urine and immediate mixing with 7 ml preservative [UCM]) leading to a final volume of 20 ml (plus or minus 2 ml). The standard urine cup collected a variable volume of first void urine and needed a manual transfer of 8 ml of urine to the 15 ml vial with 4 ml of UCM preservative by people in the study.

Key outcomes

A total of 258 urine samples from 31 people were collected and included in the analysis. A total of 12 of 33 people (61%) were detected as having high-risk HPV positivity during the study period.

The median human DNA yield in first volume urine was 7.14 (interquartile range [IQR]: 2.87 to 17.85) and 4.5 (IQR: 1.88 to 9.15) nanograms per microlitre DNA extract for the Colli‑Pee device and urine cup, respectively. Colli‑Pee collected samples had higher HPV DNA concentrations than cup-collected samples at human DNA concentrations up to 12.18 nanograms per microlitre. At higher concentrations of human DNA (more concentrated samples), the benefit of the Colli‑Pee disappeared.

Strengths and limitations

Sample size was small and included multiple samples from individuals. Study participants were self‑reported to be HPV positive. The company provided the Colli‑Pee device. Four study authors are co‑founders and board members of the company.

Leeman et al. (2017)

Intervention and comparator

People were sent Colli‑Pee to collect urine samples on the morning of colposcopy (U1). Urine samples later on (U2), brush‐based cervicovaginal self‐samples, and clinician‐taken smears were also obtained from participants.

Key outcomes

All samples were tested for HPV DNA to detect cervical intraepithelial neoplasia grade 2+ (CIN2+) SPF10 assay and the clinically validated GP5+/6+ assay.

Samples from 91 of 113 people were analysed. Sensitivity for CIN2+ with the SPF10 system using samples by clinician-taken smears, self-collected samples, U1, and U2 was 100%, 100%, 95%, and 100% respectively. With the GP5+/6+ assay, sensitivity was 95% in all sample types.

Strengths and limitations

This is a single-centre study. Strengths and limitations have not been assessed because limited information was reported in the abstract.

Sustainability

The company states that this is a single-use device made of approximately 20 g unblended polymers that are well suited for recycling. There is no published evidence to support this claim.

One expert raised the issue about throw away plasticware because it is a single-use device. The expert noted that the plastic problem is not new in cervical screening because routine screening often uses a disposable plastic speculum (rather than a reusable one that must be sterilised between uses).

Recent and ongoing studies

There are studies on the device underway with further supporting evidence being generated.