Women were excluded from the analysis if (1) they had cryotherapy performed before their initial colposcopic examination (n = 1), (2) the date of the initial colposcopic examination was not available (n = 36), (3) information confirming the type of treatment used between colposcopic examinations was unknown (n = 32), (4) initial colposcopic examination was inadequate (n = 16), or (5) the adequacy of the follow-up colposcopic examination was not documented (n = 6). The total number of women excluded was 65 because some women met multiple exclusion criteria.
The management after initial colposcopic examination was done according to whether the women had cryotherapy, cone, LEEP, or no procedure between initial and follow-up colposcopic examinations. Univariate analysis of the association between the management group with clinic of treatment, performance of ECC, biopsy results, presence of HPV, and cytologic presence of glandular atypia was performed. Mean age and duration (interval between initial and follow-up colposcopic examinations or between the procedure and follow-up examination) were calculated for all groups.
The odds ratio of an inadequate follow-up colposcopic examination was estimated for type of treatment (cryotherapy, cone/LEEP) compared with no treatment, age, clinic where treatment was provided, performance of ECC, biopsy results, presence of HPV, and presence of glandular atypia. The 95% confidence intervals about the relative odds estimates were calculated. Mean age and duration between initial colposcopy and follow-up colposcopy were calculated for the groups with adequate and inadequate follow-up colposcopic examinations.
Multivariable logistic regression analysis was used to evaluate the association of adequacy of follow-up colposcopic examination with age (years), clinic where colposcopic examination was performed, duration (days), whether or not ECC was performed, biopsy results from the initial colposcopic examination, presence of HPV, and presence of glandular atypia noted on initial colposcopy. Biopsy results were categorized as normal or abnormal in the model that is reported. The stepwise backward elimination technique was used to evaluate the best model. The 95% confidence intervals about the adjusted odds ratio were calculated.
The Pearson chi-square test was used to test the significance of the association between binary variables. The significance of the difference between means was tested with the one-way analysis of variance. Data were analyzed with the personal computer version of the Statistical Package for the Social Sciences (SPSS/PC+ version 8.0).
Results
Between January 1, 1991, and December 31, 1995, 3225 women underwent colposcopic evaluation or treatment at 7 county colposcopy clinics in Oklahoma. Two hundred sixty-eight of these women underwent 2 examinations during the study period. There were 203 of 268 subjects available for analysis after exclusions for missing data. Eighty-three patients (41.1%) had cryotherapy, 24 (11.9%) underwent a cone biopsy or a LEEP procedure, and 96 (47.5%) underwent no procedure between initial and follow-up colposcopic examinations.
Table 1 shows characteristics of women who had cryotherapy, cone/LEEP, and no procedure. The groups were similar with respect to age, whether ECC was performed, presence of HPV, and whether glandular involvement was noted. There was an association between the degree of cervical dysplasia and the three treatment groups, which was expected because degree of dysplasia determines treatment modality. Women who had cryotherapy had follow-up colposcopy (mean = 565 days) later than women who had cone or LEEP (mean = 319 days) or no procedure (mean = 339 days; P < .0001).
Thirty-three percent (n = 67) had inadequate follow-up colposcopic examinations. These included a large proportion of women, 61.4%, who had cryotherapy (51/83) compared with 20.8% (5/24) of women who had cone or LEEP and 11.5% (11/96) of women who had no procedure.
Table 2 shows the relationship between inadequate second colposcopy and previous cryotherapy, cone/LEEP, abnormal cervical biopsy, ECC, presence of HPV, and presence of glandular atypia. Patients who had cryotherapy had an increased likelihood of inadequate follow-up compared with patients who had no procedure (adjusted odds ratio =18.67, 95% confidence interval = 6.99–49.81). Cone/LEEP increased the likelihood of inadequate follow-up but was not statistically significant. Age, duration, ECC, presence of HPV, or presence of glandular atypia did not increase the likelihood of subsequent inadequate colposcopic examination. Odds ratio estimates for different clinics are not reported but were imprecise due to small numbers.
TABLE 1
Characteristics of patients with and without cryotherapy between initial and follow-up colposcopy
| Characteristic | Cryotherapy (n = 82) | Cone or LEEP (n = 24) | No procedure (n = 96) | P* |
|---|---|---|---|---|
| Mean age (y) | 24.6 | 26.5 | 23.8 | .229 |
| Mean duration†(d) | 565 | 319 | 339 | .004 |
| ECC (%) | 80.2 | 75.0 | 7.1 | .135 |
| Cervical dysplasia (%) | < .001 | |||
| Normal | 21.7 | 19.0 | 39.5 | |
| Mild dysplasia | 58.0 | 23.8 | 45.3 | |
| >Mild dysplasia | 20.3 | 57.1 | 15.1 | |
| HPV (%) | 72.6 | 52.4 | 60.4 | .130 |
| Glandular atypia (%) | 23.9 | 38.1 | 17.8 | .125 |
| *Pearson 2 for proportions and analysis of variance for means. | ||||
| †Duration from treatment (cryotherapy) or examination to follow-up colposcopic examination. | ||||
| ECC, endocervical curettage; HPV, human papilloma virus; LEEP, loop excision of the ectocervical portion. | ||||