Between 2018-2021 the STRIDES study had a population size of 32,735 individuals undergoing screening within Mississippi’s public health system, with 20,792 individuals being seen at a MSDH clinic. We identified 1,458 MSDH individuals that had an abnormal result indicating need for a colposcopy based on the 2012 ASCCP guidelines. Characteristics of these individuals by follow-up status are shown in Table 1 with a full breakdown of cytology and HPV results shown in Supplemental Table S1 and S2. A total of 627 individuals (43.0%) with an abnormal screening result had a follow-up colposcopy procedure documented within 4 months (on time). There were 239 (16.4%) individuals that had follow-up documentation of a colposcopy procedure, but outside of the recommended four-month period with a delayed mean of 15.6 months (SD = 12.91) and a maximum of 53 months seen. Of the 592 individuals (40.6%) identified as not adherent to the follow-up guidelines, 264 (18.1%) had documentation of a return visit where they received an additional co-test instead of the recommended colposcopy. Characteristics of individuals who returned for a repeat co-test compared to those who did have follow-up care documented are shown in Supplemental Table S3 .
The follow-up outcome groups had significant differences noted between adherence to follow-up recommendations by age, race/ethnicity, and cytology diagnosis shown in Table 1. Regarding age, 49.0% of individuals ≥ 30 years of age compared to 38.7% of individuals <30 years of age were adherent to follow-up guidelines (p <.001). Regarding race and ethnicity (p = .003), individuals who were Hispanic had the highest percentage (63.3%) of having follow-up care based on recommendations. Individuals listed as Other had the highest percentage (48.2%) of not having documented follow-up care. When looking at cytology diagnosis (p <.001), individuals with a low-risk cytology diagnosis such as ASC-US (53.6%) and LSIL (31.9%) were less likely to receive follow-up compared to individuals with a more severe cytology diagnosis, such as ASC-H (25.0%) and HSIL (21.6%). Individuals with an HPV-positive NILM diagnosis (44.2%) also had a higher percentage of being less likely to receive follow-up. There was no difference in follow-up outcomes by BMI category (p = 0.105) or smoking status (p = 0.216).
Table 1
Frequencies and Chi-Square Results for Follow-up Outcomes Among Individuals Following Abnormal Cervical Cancer Screening
|
Adherent: Follow-Up with Colposcopy (on time) (N=627)
|
Adherent: Follow-Up with Colposcopy (delayed) (N=239)
|
Non-Adherent: No Follow-Up (N=592)
|
P value
|
|
N
|
row%
|
N
|
row%
|
N
|
row%
|
Age
|
|
|
|
|
|
|
<.001
|
<30 years old
|
327
|
38.7%
|
126
|
14.9%
|
394
|
46.5%
|
|
≥30 years old
|
300
|
49.0%
|
113
|
18.5%
|
199
|
32.5%
|
|
Race
|
|
|
|
|
|
|
.003
|
Non-Hispanic White
|
156
|
43.7
|
57
|
16.0
|
144
|
40.3
|
|
Non-Hispanic Black
|
339
|
41.4
|
143
|
17.5
|
336
|
41.1
|
|
All Hispanic
|
57
|
63.3
|
12
|
13.3
|
21
|
23.3
|
|
Other
|
40
|
36.4
|
17
|
15.5
|
53
|
48.2
|
|
BMI
|
|
|
|
|
|
|
.105
|
<25
|
166
|
39.8
|
68
|
16.3
|
183
|
43.9
|
|
25-<30
|
125
|
38.2
|
60
|
18.3
|
142
|
43.4
|
|
30-<35
|
135
|
50.0
|
40
|
14.8
|
95
|
35.2
|
|
35+
|
147
|
43.5
|
55
|
16.3
|
136
|
40.2
|
|
Smoking
|
|
|
|
|
|
|
.216
|
Never Smoker
|
353
|
43.4
|
125
|
15.4
|
336
|
41.3
|
|
Former
|
84
|
49.1
|
29
|
17.0
|
58
|
33.9
|
|
Current
|
190
|
40.3
|
85
|
18.0
|
197
|
41.7
|
|
Cytology Diagnosis
|
|
|
|
|
|
|
<.001
|
NILMa
|
57
|
34.5
|
35
|
21.2
|
73
|
44.2
|
|
ASC-US
|
191
|
31.6
|
89
|
14.7
|
324
|
53.6
|
|
LSIL
|
212
|
52.9
|
61
|
15.2
|
128
|
31.9
|
|
ASC-H
|
37
|
51.4
|
17
|
23.6
|
18
|
25.0
|
|
HSIL
|
124
|
62.3
|
32
|
16.1
|
43
|
21.6
|
|
Note. Abbreviations: ASC-H, atypical squamous cells cannot exclude high grade; ASC-US, atypical squamous cells of undetermined significance; HSIL, high-grade intraepithelial lesion; LSIL, low-grade intraepithelial lesion; NILM, negative for intraepithelial lesion or malignancy; Pap, Papanicolaou test.
aNILM diagnoses were HPV 16, HPV 18, or HPV 12 HR other positive, indicating the need for further management with a colposcopy based on ASCCP 2012 guidelines.
We conducted multivariate logistic regression analyses to investigate factors associated with follow-up care, shown in Table 2. Overall, age (per one-year increase) was positively associated with an increased likelihood of following up with a colposcopy recommendation (OR = 1.07; 95% CI = 1.05, 1.09). When further stratified by age, younger individuals (<30 years) observed a 36% increase (OR = 1.36; 95% CI = 1.26, 1.47; p <.001) in the likelihood of following up with a colposcopy with every one-year increase in age. However, in individuals ≥30 years old, the same pattern of increased likelihood of colposcopy follow-up was not significantly identified. (OR = 1.01; 95% CI = 0.98, 1.47; p = .556). Cytology diagnoses were also associated with increased likelihood of completing a colposcopy follow-up in the various regression models. When comparing individuals with a NILM cytology result to the full sample, those with LSIL, ASC-H, and HSIL were more likely to follow up with colposcopy, (OR = 2.67, 3.9, and 5.37, respectively; p<.001). Similar findings among cytological diagnoses were observed when stratified by age groups.
Table 2
Regression Results for Follow-up with a Colposcopy Based on Total Sample (N=1458) and Age-Stratified
for < 30 (N=846) and ≥ 30 (N=612)
Variable
|
Model 1: Full Sample
|
Model 2: < 30-Year-Old
|
Model 3: ≥ 30-Year-Old
|
OR
|
95% CI
|
p
|
OR
|
95% CI
|
p
|
OR
|
95% CI
|
p
|
|
Age
|
1.07
|
[1.05,
|
1.09]
|
<.001
|
1.36
|
[1.26,
|
1.47]
|
<.001
|
1.01
|
[0.98,
|
1.04]
|
.556
|
|
Race
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Non-Hispanic White
|
Ref
|
|
|
.203
|
Ref
|
|
|
.665
|
Ref
|
|
|
.737
|
|
Non-Hispanic Black
|
1.17
|
[0.86,
|
1.57]
|
.319
|
1.13
|
[0.74,
|
1.71]
|
.579
|
1.14
|
[0.72,
|
1.80]
|
.587
|
|
All Hispanic
|
1.72
|
[0.93,
|
3.18]
|
.083
|
1.69
|
[0.68,
|
4.21]
|
.260
|
1.17
|
[0.49,
|
2.75]
|
.727
|
|
Other
|
0.88
|
[0.55,
|
1.43]
|
.611
|
0.96
|
[0.51,
|
1.81]
|
.901
|
0.75
|
[0.33,
|
1.71]
|
.490
|
|
BMI
|
|
|
|
|
|
|
|
|
|
|
|
|
<25
|
Ref
|
|
|
.312
|
Ref
|
|
|
.709
|
Ref
|
|
|
.792
|
25-<30
|
0.87
|
[0.63,
|
1.21]
|
.410
|
0.79
|
[0.51,
|
1.24]
|
.314
|
0.99
|
[0.58,
|
1.71]
|
.995
|
30-<35
|
1.19
|
[0.84,
|
1.68]
|
.338
|
1.02
|
[0.62,
|
1.66]
|
.952
|
1.27
|
[0.74,
|
2.18]
|
.390
|
35+
|
1.14
|
[0.83,
|
1.58]
|
.417
|
1.01
|
[0.65,
|
1.57]
|
.953
|
1.03
|
[0.60,
|
1.77]
|
.907
|
Smoking
|
|
|
|
|
|
|
|
|
|
|
|
|
Never Smoker
|
Ref
|
|
|
.227
|
Ref
|
|
|
.624
|
Ref
|
|
|
.064
|
Current
|
0.86
|
[0.65,
|
1.14]
|
.298
|
0.97
|
[0.66,
|
1.44]
|
.891
|
0.62
|
[0.40,
|
0.96]
|
.034
|
Former
|
1.22
|
[0.82,
|
1.81]
|
.320
|
1.28
|
[0.74,
|
2.19]
|
.378
|
1.07
|
[0.56,
|
2.05]
|
.847
|
Cytology Diagnosis
|
|
|
|
|
|
|
|
|
|
|
|
|
NILM
|
Ref
|
|
|
<.001
|
--
|
--
|
--
|
--
|
Ref
|
|
|
.005
|
ASC-US
|
1.18
|
[0.77,
|
1.81]
|
.451
|
Ref
|
|
|
<.001
|
1.59
|
[0.98,
|
2.58]
|
0.60
|
LSIL
|
2.67
|
[1.72,
|
4.14]
|
<.001
|
1.34
|
[.893,
|
2.01]
|
.157
|
2.78
|
[1.60,
|
4.81]
|
<.001
|
ASC-H
|
3.90
|
[1.91,
|
7.97]
|
<.001
|
6.60
|
[2.79,
|
15.62]
|
<.001
|
2.06
|
[0.80,
|
5.31]
|
.134
|
HSIL
|
5.37
|
[3.13,
|
9.20]
|
<.001
|
8.95
|
[4.97,
|
16.12]
|
<.001
|
2.48
|
[1.32,
|
4.68]
|
.005
|
Atypical
|
1.12
|
[0.31,
|
4.10]
|
.865
|
--
|
--
|
--
|
--
|
0.98
|
[0.26,
|
3.79]
|
.982
|
Note. Abbreviations: ASC-H, atypical squamous cells cannot exclude high grade; ASC-US, atypical squamous cells of undetermined significance; HSIL, high-grade intraepithelial lesion; LSIL, low-grade intraepithelial lesion; NILM, negative for intraepithelial lesion or malignancy; OR = Odds Ratio
Supplemental tables S1 and S2 show the distribution of cytology screening results by HPV genotype. Among the 617 individuals ≥ 30 years of age, 90.4% (n = 553) of the cytology results had an associated HPV test result, with 96.2% (n = 532) of those screening HPV positive. Among individuals <30 with ASC-US cytology diagnosis, HPV testing was completed on 99.8% (n = 428) individuals. HPV Other HR12 was the most common result among screened-positive HPV genotypes for the sample of individuals <30 (n=370, 86.2%) whereas HPV 16/18 was the most frequent subtype among HPV screened positive individuals ≥ 30 (n=281, 45.9%). Among individuals with HPV 16/18, both LSIL (OR = 4.00; 95% CI = 1.25, 12.79; p = .019) and ASC-H/HSIL (OR = 2.07, 95% CI = .93, 4.59, p = .074) cytology diagnoses results were more likely to go to colposcopy, whereas among Other HR12 only the cytology diagnosis of ASC-H/HSIL (OR = 4.16; 95% CI = 1.27, 13.67; p = .019) were more likely to receive a colposcopy for follow-up care (data not shown).
The Kaplan-Meier survival curve probability for follow-up care with a colposcopy after an abnormal screening result is displayed in Figure 1. Most individuals who underwent a colposcopy did so within 12 months (0.5 probability by 8 months), then continued to increase at a slower rate up to 0.6 at year 5. Figure 2 provides a comparison of colposcopy probability between different age groups. The curve identifies individuals aged 30 years and older had a significantly higher probability for follow-up colposcopy at nearly 0.6 after one year and 0.7 at year 5 compared to individuals aged less than 30 years (p <.001). Figure 3 represents a comparison of follow-up colposcopy probability among cytological diagnoses with significant differences noted among the five cytology diagnoses (p <.001). Individuals with a higher risk cytology diagnosis of HSIL or ACSUS had a higher probability of following up with a colposcopy (0.7 and 0.6 after one year, respectively) compared to other diagnoses.