Cervical cancer progression and regression data from a retrospective cohort study at Meru Level 5 Hospital, Kenya (2018–2023)

This dataset contains de‑identified individual‑level data from 550 women who underwent cervical cancer screening and follow‑up at Meru Level 5 Hospital, Kenya, between January 1, 2018 and December 31, 2023. The data were used to perform a competing risks survival analysis of cervical lesion progression and regression.

Variables included:

• Age – Age at baseline screening (years)

• HPV_Status – High‑risk HPV infection status (0 = negative, 1 = positive)

• HPV_Type_HighRisk – High‑risk HPV genotype (0 = none, 1 = HPV16, 2 = HPV18, 3 = other)

• Smoking – Self‑reported smoking (0 = never, 1 = current/former)

• Parity – Number of full‑term pregnancies

• Contraceptive_Use – Current use of hormonal contraceptives (0 = no, 1 = yes)

• HIV_Status – HIV serostatus (0 = negative, 1 = positive)

• Immune_Status – Immune competence category (1 = normal, 2 = moderate suppression, 3 = severe suppression)

• Screening_Frequency – Number of cervical screening visits in the previous five years (0, 1, 2)

• CIN_Stage – Baseline cervical intraepithelial neoplasia grade (0 = no lesion, 1 = CIN1, 2 = CIN2, 3 = CIN3)

• Cancer_Status – Invasive cervical cancer diagnosis during follow‑up (0 = no, 1 = yes)

• Time_in_State – Follow‑up time in years from baseline to last contact, event, or censoring

• Progression_Event – Binary indicator of progression to a higher‑grade lesion (1 = yes)

• Regression_Event – Binary indicator of regression to a lower‑grade lesion or normal histology (1 = yes)

Study population:

Women aged ≥18 years with documented cervical histology or HPV results and complete follow‑up information. Exclusion criteria included prior invasive cervical cancer, hysterectomy, pregnancy at screening, or incomplete records.

Data use and access:

The dataset is fully anonymised and contains no personally identifiable information. It is provided under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence to allow reuse with proper attribution. The data accompany the manuscript submitted to PLOS ONE, and the code for the analysis is available in the supplementary materials.

File format:
Microsoft Excel (.xlsx)

Rows / columns:
550 rows (patients) × 14 columns (variables)