Table 3.

Risk of bias in time-trend analyses

First author (year)	Risk of selection bias: changes in the study population characteristics between the pre- and post-vaccination periods	Risk of information bias: errors in the identification of HPV+ during the pre- and post-vaccination period (data source, genital wart case definition, outcome used)	Risk of confounding: changes in HPV infection between the pre- and post-vaccination periods could be diluted/exacerbated by other variables
Dominiak-Felden (2015) [17]	High Some people possibly may have been vaccinated without reimbursement (risk of misclassification), imiquimod agreement was used as the date of vaccination.	High A surrogate marker (imiquimod agreement) was used as definition of genital wart cases (risk of underestimation).	High Different sexual behaviours between vaccinated and unvaccinated women.
Chow (2015) [28]	High Possible changes in the clientele of the sexual health services between the periods. Self-reported vaccination status and the number of doses of HPV vaccine.	Low Clinical diagnosis by clinicians.	High Clients at sexual health service have higher risk of sexually transmissible infections.
Ali (2013) [27]	High Possible changes in the clientele of the sexual health services between the periods. Self-reported vaccination status and the number of doses of HPV vaccine.	Low Genital warts are directly diagnosed by physicians.	High Changes in sexual activity, health seeking behaviour could potentially cause changes in genital wart frequency over time.
Harrison (2014) [29]	Unclear Some women from the vaccination eligible group may be included with non-vaccine eligible women due to the change in patient age.	Low Genital wart diagnosis by physicians.	High Change in sexual risk behaviour.
Read (2011) [31]	Unclear Possible changes in the clientele of the Melbourne Sexual Health Centre.	Low Genital wart diagnosis by physicians.	Unclear Possible HPV infection of 21–29 years women before the vaccination.
Fairley (2009) [30]	Unclear Possible changes in the clientele of the Melbourne Sexual Health Centre.	Low Genital wart diagnosis by physicians.	Unclear Boys aged 9–15 years could be prescribed the vaccine privately.
Checchi (2019) [32]	Unclear Inability to link anogenital wart diagnoses to individual vaccination status.	Low Genital wart diagnosis by physicians.	Unclear Patients could attend elsewhere for treatment of anogenital wart.
Mann (2019) [33]	Unclear Patients’ vaccination status is unknown.	Low Genital wart diagnosis by physicians.	Unclear Patients with anogenital wart could choose to seek care elsewhere.