Table 1 Participant demographics for 15 included studies
| Â | Study design | Country | Inclusion criteria | Exclusion criteria | Type of colposcope | Sample size | Time of recruitment | Number and type of biopsies | Referral cytology/HPV testing | The level of colposcopists | Mean age (SD; range) | Main Conclusion |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Ghosh et al., 2014 [24] | Retrospective | India | Intact uterus; without a history of precancer/cancer; nonpregnant | NR | NR | 2466 | 2011–2013 |  ≥ 1; abnormal areas | HC2 | ≥ 10 years | NR (NR; 30–60) | Colposcopy performed well in the overall detection of cervical neoplasias, though its capability for accurate categorisation of degree of abnormality was poor |
Spinillo et al., 2014 [28] | Retrospective | Italy | NR | Pregnancy, treatment, hysterectomy; | NR | 2526 | 2009–2012 | 2–4; abnormal areas and random biopsies | Thinprep cytology; HPV | NR | 37 (29–45; NR) | Multiple infection or HPV16 positivity did not affect colposcopic accuracy in the diagnosis of CIN3+ lesions. The sensitivity of colposcopy was poor among subjects who were uninfected or infected by low-risk HPV genotypes |
Zhao et al., 2015 [30] | Retrospective | China | No history of hysterectomy, pregnancy, pelvic radiotherapy, screening | NR | NR | 1997 | 1999 | 2–4; abnormal areas and random biopsies | Thinprep cytology; HC2 | NR | 39.6 (3.2;35–45) | 4-quadrant biopsy can detect more HSIL+ lesions and is more accurate than suspicious lesion biopsy alone |
Coronado et al., 2016 [22] | Retrospective | Spain | Aged ≥ 18 years | NR | Digital colposcopy | 443 | 2012–2014 | 2; abnormal areas | Pap smear; HPV | Accredited as experts by the SSCPC | 36.0 (10.9; NR) | Combining conventional colposcopy with DSI mapping improves the capability to detect cervical lesions |
Li et al., 2017 [8] | Retrospective | China | NR | Had a history of hysterectomy or treatment; incomplete data | Digital colposcopy | 525 | 2014–2015 |  ≥ 1;abnormal areas and random biopsies | HC2 or Cobas HPV | Received training by 2011 IFCPC | 40.13 (10.23; NR) | The 2011 IFCPC nomenclature improves colposcopic accuracy in trained colposcopists, like speaking the same language. However, the reproducibility of TZ and the predictive value of a few signs remain to be discussed |
Fan et al., 2018 [9] | Retrospective | China | No treatment history hysterectomy; no clinically suspected immunosuppression | NR | Digital colposcopy | 2262 | 2012–2016 |  ≥ 1;abnormal areas and random biopsies | Cytology; HR-HPV | ≥ 5 years | 41.3 (11.6; NR) | The 2011 IFCPC terminology can improve the diagnostic accuracy for all lesion severities. The categorization of major changes and minor changes is appropriate. However, colposcopic diagnosis remains unsatisfactory |
Liu et al., 2018 [31] | Retrospective | China | Complete case data | Patients with a history of pathology or surgery and pelvic radiotherapy | NR | 256 | 2014–2016 |  ≥ 1; abnormal areas and random biopsies | Pap smear; NR | Senior | 47 (NR;23–80) | The type of transformation zone is positively correlated with the age, and it can help to choose biopsy and therapeutic manner. The diagnostic accuracies of HSIL and early stage of cervical cancer by multi-point biopsy of colposcopy and/or ECC are high |
Ruan et al., 2020 [27] | Retrospective | China | NR | Unclear image or the patients lacked the image; no HPV, cytology or histopathology | Digital colposcopy | 1828 | 2016–2019 | NR | Cytology; HPV | ≥ 20 years | 37 (NR; 17–81) | The data and findings herein provide the resource for evaluating the diagnostic value of colposcopy, and suggested that the accuracy of colposcopy is required to be further improved |
Del Pino et al., 2021 [23] | Prospective | Spain | NR | Previous treatment; pregnancy; | Digital colposcopy | 302 | 2014–2015 | 2–4; abnormal areas and random biopsies | ThinPrep cytology; Cobas HPV | accredited by the SSCPC | 37.6 (10.3; NR) | Colposcopy impression provides essential information to identify women at risk of HSIL/CIN3 |
Li et al., 2021 [25] | Retrospective | China | NR | No screening results, had uterectom or CIN; pregnant or with incomplete data | NR | 495 | 2017–2019 |  ≥ 1;abnormal areas and random biopsies | ThinPrep cytology; HC2, genotype | NR | 40 (NR; 21–71) | Colposcopy is an excellent tool to estimate cervical high-grade lesion but is imprecise. Many factors can bias the diagnosis of colposcopy, especially the known results of cervical cytology and HPV |
Liu et al., 2021 [26] | Retrospective | China | NR | Pregnancy; previous lesions or surgery, hysterectomy; incomplete data | Digital colposcopy | 987 | 2015–2019 | ≥ 1;abnormal areas | Cytology | NR | 41.94 (12.45; NR) | The diagnostic value of IFCPC and R-way is better than Reid. There is good agreement between R-way colposcopy evaluation and histopathology |
Zhang et al., 2022 [33] | Retrospective | China | NR | Had hysterectomy or pelvic radiation; no histopathology | Digital colposcopy | 1838 | 2013–2018 | 2–4; abnormal areas and random biopsies | Cytology; HR-HPV | 5–7 years | 41.7 (10.6; NR) | Positive p16(INK4a) immunostaining is very strongly consistent with an H&E diagnosis of CIN2+, and it can be used as an objective detection index for HSIL + diagnoses of HPV-negative patients with CIN2+ |
Maffini et al., 2022 [34] | Retrospective | Brazil | Aged ≥ 21 years; no treatment; no history of hysterectomy | No colposcopic records; inadequate cytology review | NR | 102 | 2009–2016 | Biopsy the worst area | Pap smear; NR | 30 years | 36 (NR; 21–84) | Colposcopy performed by an experienced examiner can accurately differentiate patients with CIN1 or less from patients with CIN2 or worse. Diagnosis of CIN2 or worse was more frequent in patients with a previous history of cervical dysplasia and pre-menopausal patients. The degree of acetowhite changes was the best colposcopic feature to predict CIN2 or worse |
Wei et al., 2022 [29] | Retrospective | China | NR | Had treatment or hysterectomy; had no histologic report | Digital colposcopy | 2417 | 2018–2021 | Abnormal areas and random biopsies | NR | Senior and junior | NR | It appears possible to supplement colposcopic examinations with screening results to improve HSIL+ detection, especially for women with TZ3 lesions. It may also be possible to improve junior colposcopists' diagnoses although, further psychological research is necessary |
Stuebs et al., 2022 [32] | Retrospective | Erlangen | Had a biopsy or underwent excisional surgical treatment—LLETZ, loop electrosurgical excision procedure with laser coagulation of the periphery or laser conization | Without a biopsy being taken during the colposcopic examination | Digital colposcopy | 4778 | 2015–2022 | Biopsy the worst area | Pap smear; NR | Experienced and highly qualified | 36.8(10.8, NR) | Colposcopy is an important, feasible, and effective method. Careful work-up needs to be performed for women with TZ3 who are over 35 years old, as they are at the highest risk of being misdiagnosed. The highest concordance for detecting HSIL + was seen for colposcopists with > 10 years’ experience |