Determinants of Cancer Screening in India: An Epidemiological Overview
Dharmendra Kumar Dubey, Pramod Mishra
TL;DR
This study examines cancer screening rates and their determinants in India, finding very low coverage and identifying factors that influence screening behavior.
Contribution
The study identifies sociodemographic and behavioral determinants of cancer screening in India using nationally representative data.
Findings
Only 1.2% of women in India reported cervical cancer screening, with similarly low rates for breast and oral cancers.
Exposure to health messages via TV and radio, higher BMI, and spouse's education were linked to increased screening.
High parity, rural residence, and breastfeeding were associated with lower screening uptake.
Abstract
Background: Noncommunicable diseases (NCDs) have emerged as a major global health burden. Among these, cancer is the second leading cause of mortality worldwide and accounts for a substantial proportion of deaths in India. This study aimed to evaluate state-specific disparities in breast, cervical, and oral cancer screening coverage and to identify determinants of cancer screening patterns. Methods and materials: Data relevant to the study objectives were obtained from the fifth National Family Health Survey, which follows a cross-sectional study design. For analysis, only women aged 15-49 years were included. The primary outcomes of interest were self-reported screening behaviors for cervical, breast, and oral cancers. Responses were dichotomized as “yes” (screened) or “no” (not screened). Three separate models, each focusing on a distinct set of independent variables, were used to…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
| State | Total number of women interviewed (N) | Ever screened for cervical cancer, n (%) | Ever screened for breast cancer, n (%) | Ever screened for oral cancer, n (%) |
| Jammu and Kashmir | 1,620 | 6 (0.4%) | 4 (0.2%) | 11 (0.7%) |
| Himachal Pradesh | 951 | 5 (0.5%) | 4 (0.4%) | 6 (0.6%) |
| Punjab | 3,906 | 149 (3.8%) | 21 (0.5%) | 20 (0.5%) |
| Chandigarh | 136 | 5 (3.7%) | 0 (0.0%) | 0 (0.0%) |
| Uttarakhand | 1,753 | 4 (0.2%) | 3 (0.2%) | 7 (0.4%) |
| Haryana | 4,300 | 16 (0.4%) | 13 (0.3%) | 12 (0.3%) |
| NCT of Delhi | 2,822 | 6 (0.2%) | 3 (0.1%) | 13 (0.5%) |
| Rajasthan | 14,730 | 36 (0.2%) | 22 (0.1%) | 24 (0.2%) |
| Uttar Pradesh | 44,190 | 403 (0.9%) | 122 (0.3%) | 225 (0.5%) |
| Bihar | 29,918 | 113 (0.4%) | 89 (0.3%) | 124 (0.4%) |
| Sikkim | 58 | 1 (1.7%) | 0 (0.0%) | 0 (0.0%) |
| Arunachal Pradesh | 171 | 1 (0.6%) | 1 (0.6%) | 1 (0.6%) |
| Nagaland | 213 | 0 (0.0%) | 0 (0.0%) | 1 (0.5%) |
| Manipur | 448 | 5 (1.1%) | 5 (1.1%) | 3 (0.7%) |
| Mizoram | 176 | 7 (4.0%) | 2 (1.1%) | 1 (0.6%) |
| Tripura | 598 | 2 (0.3%) | 1 (0.2%) | 3 (0.5%) |
| Meghalaya | 856 | 7 (0.8%) | 3 (0.4%) | 2 (0.2%) |
| Assam | 5,791 | 11 (0.2%) | 10 (0.2%) | 12 (0.2%) |
| West Bengal | 15,542 | 14 (0.090%) | 15 (0.096%) | 12 (0.077%) |
| Jharkhand | 6,847 | 30 (0.4%) | 10 (0.1%) | 14 (0.2%) |
| Odisha | 7,112 | 48 (0.7%) | 8 (0.1%) | 10 (0.1%) |
| Chhattisgarh | 4,991 | 20 (0.4%) | 7 (0.1%) | 7 (0.1%) |
| Madhya Pradesh | 13,424 | 151 (1.1%) | 117 (0.9%) | 147 (1.1%) |
| Gujarat | 9,610 | 24 (0.2%) | 8 (0.1%) | 16 (0.2%) |
| Dadra and Nagar Haveli and Daman and Diu | 87 | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) |
| Maharashtra | 17,905 | 405 (2.3%) | 276 (1.5%) | 336 (1.9%) |
| Andhra Pradesh | 7,210 | 219 (3.0%) | 34 (0.5%) | 220 (3.1%) |
| Karnataka | 9,909 | 43 (0.4%) | 14 (0.1%) | 20 (0.2%) |
| Goa | 200 | 2 (1.0%) | 2 (1.0%) | 3 (1.5%) |
| Lakshadweep | 10 | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) |
| Kerala | 4,412 | 59 (1.3%) | 46 (1.0%) | 13 (0.3%) |
| Tamil Nadu | 10,603 | 792 (7.5%) | 375 (3.5%) | 95 (0.9%) |
| Puducherry | 141 | 8 (5.7%) | 4 (2.9%) | 1 (0.7%) |
| Andaman and Nicobar Islands | 39 | 1 (2.6%) | 1 (2.6%) | 4 (10.3%) |
| Telangana | 4,861 | 49 (1.01%) | 10 (0.2%) | 58 (1.2%) |
| Ladakh | 27 | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) |
| India | 225,565 | 2,642 (1.2%) | 1,230 (0.5%) | 1,421 (0.6%) |
| Associates | Ever undergone screening for cervical cancer n (%) | Ever undergone breast cancer screening n (%) | Ever undergone a screening for oral cancer n (%) | Total |
| Sociodemographic determinants | ||||
| Place of residence | ||||
| Urban | 947 (1.6) | 518 (0.9) | 541 (0.9) | 59,204 |
| Rural | 1,696 (1.0) | 712 (0.4) | 880 (0.5) | 166,363 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,567 |
| Age | ||||
| <20 | 51 (0.8) | 17 (0.3) | 22 (0.4) | 6,001 |
| 20-34 | 2,321 (1.2) | 1,099 (0.5) | 1,283 (0.6) | 200,970 |
| 35-49 | 271 (1.5) | 114 (0.6) | 116 (0.6) | 18,596 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,567 |
| Highest education level | ||||
| No education | 260 (0.5) | 125 (0.3) | 203 (0.4) | 48,303 |
| Primary | 239 (0.9) | 77 (0.3) | 137 (0.5) | 27,883 |
| Secondary | 1,390 (1.2) | 598 (0.5) | 765 (0.7) | 114,621 |
| Higher | 754 (2.2) | 430 (1.2) | 315 (0.9) | 34,760 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,420 (0.6) | 225,567 |
| Caste | ||||
| Schedule caste | 708 (1.3) | 267 (0.5) | 298 (0.6) | 52,718 |
| Schedule tribe | 139 (0.6) | 64 (0.3) | 98 (0.4) | 22,790 |
| Other backward class | 1,339 (1.4) | 646 (0.7) | 650 (0.7) | 98,088 |
| None of them | 350 (0.9) | 164 (0.4) | 287 (0.7) | 39,792 |
| Don't know | 6 (0.3) | 8 (0.4) | 2 (0.1) | 2,087 |
| Total | 2,542 (1.2) | 1,149 (0.5) | 1,335 (0.6) | 215,475 |
| Religion | ||||
| Hindu | 2,133 (1.2) | 1,001 (0.6) | 1,143 (0.6) | 179,506 |
| Muslim | 232 (0.6) | 123 (0.3) | 187 (0.5) | 36,278 |
| Christian | 132 (2.8) | 60 (1.3) | 42 (0.9) | 4,701 |
| Sikh | 108 (3.9) | 24 (0.9) | 25 (0.9) | 2,769 |
| Buddhist/neo-Buddhist | 21 (1.9) | 16 (1.5) | 17 (1.6) | 1,085 |
| Jain | 14 (4.7) | 4 (1.3) | 4 (1.3) | 300 |
| Others (Jewish, Parsi, no religion, and other) | 2 (0.2) | 2 (0.2) | 2 (0.2) | 927 |
| Total | 2,642 (1.2) | 1,230 (0.5) | 1,420 (0.6) | 225,566 |
| Wealth index | ||||
| Poorest | 345 (0.6) | 164 (0.3) | 266 (0.5) | 55,841 |
| Poorer | 415 (0.8) | 183 (0.4) | 225 (0.5) | 49,270 |
| Middle | 575 (1.3) | 266 (0.6) | 275 (0.6) | 44,277 |
| Richer | 661 (1.6) | 274 (0.4) | 315 (0.8) | 41,449 |
| Richest | 646 (1.9) | 343 (1.0) | 341 (1.0) | 34,729 |
| Total | 2,642 (1.2) | 1,230 (0.5) | 1,422 (0.6) | 225,566 |
| Reproductive health determinants | ||||
| Birth order | ||||
| 1st | 1,134 (1.3) | 541 (0.6) | 613 (0.7) | 88,107 |
| 2nd and 3rd | 1,290 (1.2) | 582 (0.5) | 677 (0.6) | 110,383 |
| 4th and 5th | 180 (0.8) | 89 (0.4) | 110 (0.5) | 21,738 |
| 6th + | 39 (0.7) | 18 (0.3) | 21 (0.4) | 5,340 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,568 |
| Currently breastfeeding | ||||
| No | 1,229 (1.5) | 597 (0.7) | 560 (0.7) | 80,993 |
| Yes | 1,414 (1.0) | 633 (0.4) | 861 (0.6) | 144,575 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,568 |
| Ever had a terminated pregnancy | ||||
| No | 2,186 (1.2) | 1,018 (0.5) | 1,221 (0.6) | 190,017 |
| Yes | 457 (1.3) | 212 (0.6) | 200 (0.6) | 35,550 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,567 |
| Current marital status | ||||
| Never in union | 0 (0.0) | 0 (0.0) | 1 (0.5) | 221 |
| Married | 2,621 (1.2) | 1,213 (0.5) | 1,410 (0.6) | 222,983 |
| Widowed | 6 (0.5) | 4 (0.3) | 1 (0.1) | 1,276 |
| Divorced | 0 (0.0) | 0 (0.0) | 1 (0.5) | 191 |
| No longer living together/separated | 15 (1.7) | 12 (1.3) | 9 (1.0) | 895 |
| Total | 2,642 (1.2) | 1,229 (0.5) | 1,422 (0.6) | 225,566 |
| Currently residing with husband/partner | ||||
| Living with her | 2,335 (1.2) | 1,079 (0.6) | 1,282 (0.7) | 194,584 |
| Staying elsewhere | 286 (1.0) | 134 (0.5) | 128 (0.5) | 286 |
| Total | 2,621 (1.2) | 1,213 (0.5 ) | 1,410 (0.6) | 222,982 |
| Contraceptive use and intention | ||||
| Using modern methods | 1,329 (1.3) | 550 (0.5) | 626 (0.6) | 103,317 |
| Using traditional methods | 180 (0.6) | 78 (0.3) | 86 (0.3) | 292,08 |
| Non-user-intends to use later | 254 (1.5) | 141 (0.8) | 156 (0.9) | 17,380 |
| Does not intend to use | 880 (1.2) | 462 (0.6) | 553 (0.7) | 75,615 |
| Never had sex | 0 (0.0) | 0 (0.0) | 0 (0.0) | 46 |
| Total | 2,643 (1.2) | 1,231 (0.5) | 1,421 (0.6) | 225,566 |
| Ancillary factors | ||||
| Working status | ||||
| No | 362 (1.3) | 145 (0.5) | 142 (0.5) | 28,499 |
| Yes | 80 (1.4) | 34 (0.6) | 28 (0.5) | 5,851 |
| Total | 442 (1.2) | 179 (0.5) | 170 (0.5) | 34,350 |
| Body mass index | ||||
| Underweight | 387 (0.9) | 166 (0.4) | 210 (0.5) | 43,313 |
| Healthy weight | 1,403 (1.0) | 644 (0.5) | 722 (0.5) | 137,025 |
| Overweight | 588 (1.8) | 276 (0.8) | 250 (0.8) | 32,587 |
| Obesity | 235 (2.4) | 119 (1.2) | 102 (1.0) | 9,933 |
| Total | 441 (0.3) | 180 (0.5) | 170 (0.5) | 34,316 |
| Husband/partner education | ||||
| No education | 33 (0.6) | 10 (0.2) | 17 (0.3) | 5,227 |
| Primary | 65 (1.4) | 13 (0.3) | 20 (0.4) | 4,570 |
| Secondary | 222 (1.2) | 100 (0.5) | 91 (0.5) | 18,730 |
| Higher | 121 (2.1) | 57 (1.0) | 42 (0.7) | 5,639 |
| Don't know | 0 (0.0) | 0 (0.0) | 0 (0.0) | 150 |
| Total | 2,613 (1.2) | 1,205 (0.5) | 1,284 (0.6) | 222,858 |
| Heard family planning on the radio for the last few months | ||||
| No | 2,100 (1.1) | 879 (0.5) | 1,081 (0.6) | 191,735 |
| Yes | 542 (1.6) | 351 (1.0) | 340 (1.0) | 33,831 |
| Total | 2,642 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,566 |
| Heard family planning on TV for the last few months | ||||
| No | 862 (0.6) | 403 (0.4) | 554 (0.6) | 98,303 |
| Yes | 1,781 (1.4) | 827 (0.6) | 866 (0.7) | 127,265 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,420 (0.6) | 225,568 |
| Heard about family planning in the newspaper/magazine for the last few months | ||||
| No | 1,557 (1.0) | 701 (0.5) | 850 (0.6) | 149,413 |
| Yes | 1,086 (1.4) | 529 (0.7) | 571 (0.7) | 76,154 |
| Total | 2,643 (1.2) | 1,230 (0.5) | 1,421 (0.6) | 225,567 |
| Predictors | Ever undergone a screening for cervical cancer OR (p) (95% CI) | Ever undergone breast cancer screening OR (p) (95% CI) | Ever undergone a screening for oral cancer OR (p) (95% CI) |
| Model 1: sociodemographic determinants | |||
| Place of residence | p = 0.00 | p = 0.00 | p = 0.00 |
| Urban | Ref | Ref | Ref |
| Rural | 0.87 (0.004) (0.79-0.96) | 0.68 (0.00) (0.59-0.78) | 0.74 (0.004) (0.65-0.84) |
| Age | p = 0.00 | p = 0.02 | p = 0.28 |
| Less than 20 | Ref | Ref | Ref |
| 20-34 | 1.13 (0.40) (0.85-1.52) | 1.33 (0.25) (0.82-2.15) | 1.40 (0.12) (0.92-2.15) |
| 35-49 | 1.59 (0.04) (1.16-2.18) | 1.72 (0.04) (1.02-2.87) | 1.45 (0.12) (0.91-2.31) |
| Highest education level | p = 0.00 | p = 0.00 | p = 0.001 |
| No education | Ref | Ref | Ref |
| Primary | 1.52 (0.00) (1.27-1.81) | 1.03 (0.83) (0.77-1.38) | 1.13 (0.28) (0.94-1.41) |
| Secondary | 1.83 (0.00) (1.58-2.11) | 1.67 (0.00) (1.35-2.05) | 1.35 (0.00) (1.14-1.61) |
| Higher | 2.86 (0.00) (2.43-3.37) | 3.48 (0.00) (2.75-4.39) | 1.49 (0.00) (1.21-1.85) |
| Caste | p = 0.00 | p = 0.00 | p = 0.00 |
| Schedule caste | Ref | Ref | Ref |
| Schedule tribe | 0.52 (0.00) (0.43-0.63) | 0.64 (0.002) (0.48-0.84) | 0.85 (0.17) (0.67-1.07) |
| Other backward class | 1.02 (0.66) (0.93-1.13) | 1.28 (0.001) (1.09-1.48) | 1.21 (0.008) (1.05-1.40) |
| None of them | 0.55 (0.00) (0.48-0.63) | 0.67 (0.00) (0.55-0.82) | 1.23 (0.02) (1.03-1.46) |
| Religion | p = 0.00 | p = 0.00 | p = 0.00 |
| Hindu | Ref | Ref | Ref |
| Muslim | 0.58 (0.00) (0.49-0.68) | 0.54 (0.00) (0.43-0.68) | 0.66 (0.00) (0.55-0.79) |
| Christian | 2.30 (0.00) (1.90-2.79) | 1.97 (0.00) (1.47-2.64) | 1.17 (0.38) (0.82-1.67) |
| Sikh | 2.88 (0.00) (2.33-3.56) | 1.25 (0.37) (0.77-2.00) | 0.99 (0.96) (0.62-1.58) |
| Buddhist/neo-Buddhist | 1.36 (0.17) (0.88-2.11) | 2.51 (0.00) (1.49-4.21) | 2.70 (0.00) (1.66-4.40) |
| Jain | 0.99 (0.98) (0.35-2.77) | 1.76 (0.29) (0.62-4.94) | 1.51 (0.44) (0.54-4.23) |
| Others (Jewish, Parsi, no religion, and other) | 0.38 (0.19) (0.08-1.64) | 0.94 (0.93) (0.25-3.61) | 0.58 (0.41) (0.15-2.18) |
| Wealth index | p = 0.00 | p = 0.06 | p = 0.09 |
| Poorest | Ref | Ref | Ref |
| Poorer | 1.15 (0.06) (0.99-1.34) | 1.03 (0.79) (0.83-1.28) | 0.83 (0.05) (0.69-1.00) |
| Middle | 1.51 (0.00) (1.30-1.74) | 1.30 (0.02) (1.05-1.61) | 0.96 (0.66) (0.78-1.15) |
| Richer | 1.59 (0.00) (1.38-1.86) | 1.08 (0.49) (0.86-1.36) | 1.03 (0.74) (0.85-1.26) |
| Richest | 1.49 (0.00) (1.27-1.78) | 1.15 (0.26) (0.89-1.48) | 1.09 (0.43) (0.88-1.36) |
| Model-2: reproductive health determinants | |||
| Birth order | p = 0.00 | p = 0.00 | p = 0.001 |
| 1st order | Ref | Ref | Ref |
| 2nd and 3rd order | 0.88 (0.00) (0.80-0.95) | 0.82 (0.00) (0.73-0.93) | 0.88 (0.02) (0.79-0.98) |
| 4th and 5th order | 0.63 (0.00) (0.54-0.74) | 0.66 (0.00) (0.53-0.82) | 0.74 (0.00) (0.60-0.90) |
| 6+ order | 0.55 (0.00) (0.40-0.76) | 0.52 (0.00) (0.32-0.84) | 0.56 (0.00) (0.36-0.86) |
| Currently breastfeeding | p = 0.00 | p = 0.00 | p = 0.002 |
| No | Ref | Ref | Ref |
| Yes | 0.63 (0.00) (0.59-0.69) | 0.59 (0.00) (0.53-0.66) | 0.84 (0.00) (0.76-0.94) |
| Ever had a terminated pregnancy | p = 0.052 | p = 0.22 | p = 0.100 |
| No | Ref | Ref | Ref |
| Yes | 1.11 (0.05) (0.99-1.23) | 1.09 (0.22) (0.94-1.27) | 0.88 (0.10) (0.76-1.03) |
| Currently residing with husband/partner | p = 0.016 | p = 0.15 | p = 0.15 |
| Living with her | Ref | Ref | Ref |
| Staying elsewhere | 0.86 (0.01) (0.76-0.97) | 0.88 (0.15) (0.73-1.05) | 0.69 (0.00) (0.57-0.83) |
| Model 3: ancillary factors | |||
| Working status | p = 0.529 | p = 0.46 | p = 0.969 |
| No | Ref | Ref | Ref |
| Yes | 1.08 (0.52) (0.85-1.38) | 1.15 (0.46) (0.79-1.68) | 1.01 (0.97) (0.67-1.52) |
| Body mass index | p = 0.00 | p = 0.00 | p = 0.028 |
| Underweight | Ref | Ref | Ref |
| Healthy weight | 0.92 (0.56) (0.71-1.21) | 1.35 (0.24) (0.82-2.23) | 1.12 (0.62) (0.71-1.78) |
| Overweight | 1.78 (0.00) (1.32-2.40) | 3.24 (0.00) (1.90-5.51) | 1.86 (0.02) (1.10-3.15) |
| Obesity | 1.04 (0.87) (0.64-1.69) | 2.74 (0.00) (1.39-5.41) | 1.78 (0.11) (0.88-3.61) |
| Contraceptive use and intention | p = 0.015 | p = 0.165 | p = 0.271 |
| Using modern methods | Ref | Ref | Ref |
| Using traditional methods | 0.59 (0.00) (0.43-0.84) | 0.76 (0.27) (0.47-1.24) | 1.16 (0.57) (0.70-1.90) |
| Non-user-intends to use later | 0.91 (0.59) (0.64-1.29) | 1.21 (0.46) (0.73-1.98) | 1.78 (0.02) (1.07-2.98) |
| Does not intend to use | 0.76 (0.01) (0.61-0.94) | 0.67 (0.03) (0.46-0.98) | 1.24 (0.24) (0.87-1.79) |
| Never had sex | 0.00 (1.00) (0.00-0.00) | 0.00 (1.00) (0.00-0.00) | 0.00 (1.00) (0.00-0.00) |
| Heard family planning on the radio for the last few months | p = 0.069 | p = 0.00 | p = 0.00 |
| No | Ref | Ref | Ref |
| Yes | 1.26 (0.06) (0.98-1.60) | 2.35 (0.00) (1.68-3.29) | 2.47 (0.00) (1.74-3.51) |
| Heard family planning on TV for the last few months | p = 0.00 | p = 0.00 | p = 0.017 |
| No | Ref | Ref | Ref |
| Yes | 1.63 (0.00) (1.29-2.07) | 2.09 (0.00) (1.39-3.11) | 1.66 (0.01) (1.09-2.51) |
| Heard about family planning in the newspaper/magazine for the last few months | p = 0.00 | p = 0.00 | p = 0.326 |
| No | Ref | Ref | Ref |
| Yes | 0.99 (0.95) (0.79-1.24) | 0.72 (0.06) (0.51-1.01) | 1.20 (0.32) (0.83-1.73) |
| Husband/partner education | p = 0.00 | p = 0.001 | p = 0.788 |
| No education | Ref | Ref | Ref |
| Primary | 2.06 (0.00) (1.35-3.14) | 1.44 (0.39) (0.63-3.29) | 1.22 (0.55) (0.64-2.34) |
| Secondary | 1.53 (0.02) (1.06-2.22) | 2.31 (0.01) (1.18-4.51) | 1.08 (0.79) (0.63-1.84) |
| Higher | 2.50 (0.00) (1.68-3.74) | 3.61 (0.00) (1.78-7.31) | 1.35 (0.32) (0.74-2.46) |
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsGlobal Cancer Incidence and Screening · Cervical Cancer and HPV Research · Colorectal Cancer Screening and Detection
Introduction
Noncommunicable diseases (NCDs) have emerged as a pervasive global health burden. In 2021, an estimated 43 million deaths were attributed to NCDs, of which 18 million occurred prematurely before the age of 70. These deaths accounted for over 75% of all non-pandemic-related mortality worldwide. Alarmingly, 82% of these deaths occurred in low- and middle-income countries [1].
In 2022, cancer was the second leading contributor to global mortality among NCDs. Nearly half of all individuals diagnosed with cancer succumbed to the disease in the same year, with 9.7 million deaths reported among 20 million new cases worldwide [2]. Asia, home to the largest share of the global population, bore a disproportionate burden, accounting for 49.2% of new cancer cases and 56.1% of cancer-related deaths globally.
Breast cancer emerged as the most frequently diagnosed malignancy, accounting for 2.2 million new cases (11.3% of all cancer diagnoses) and 7.0% of cancer-related deaths worldwide. Cervical cancer, another major public health concern, accounted for 6.0 million new cases (3.0%) and 3.4 million deaths (3.5%). Oral cancer contributed 3.8 million new cases (1.9%) and 1.8 million deaths (1.8%) globally [3,4].
According to the World Health Organization NCD Country Profile for India, the national burden of NCDs remains substantial. NCDs account for approximately 63% of all deaths in India, with cancers contributing nearly 9% of total NCD-related mortality [5,6]. In 2022, India recorded approximately 917,000 cancer-related deaths among 1.41 million individuals (0.1% of the population) diagnosed with cancer. Among women, breast cancer emerged as the leading cause of cancer-related mortality, with 190,000 new cases (13.6%) and 98,337 deaths. Cervical cancer followed, with 130,000 cases (9.0%) and 79,906 deaths (8.7%). Predominantly affecting men, oral cancer accounted for 143,000 new cases (10.2%) and 79,979 deaths (8.7%), ranking second overall among all cancers in the same year [2].
Projections from the Indian Council of Medical Research-National Cancer Registry Programme indicate a concerning rise in India’s cancer burden, from 1.4 million cases in 2022 to an estimated 1.57 million cases by 2025 [7]. Despite the presence of public health schemes, approximately 75% of cancer-related healthcare expenditures are incurred out of pocket, with cancer care costs estimated to be nearly six times higher than those for infectious diseases [8-12]. The escalating incidence of cancer thus imposes a substantial financial burden on affected households [13]. Notably, cancer-related hospitalizations carry a 160% higher risk of catastrophic health expenditure compared with communicable diseases and incur nearly twice the costs associated with injuries or cardiovascular conditions [14].
Screening for oral, breast, and cervical cancers has been shown to substantially improve five-year survival rates, increasing to nearly 60% when cancers are detected at an early stage but declining to below 15% when diagnosed at advanced stages [15]. Accordingly, screening programs aim to detect early, asymptomatic, and precancerous lesions [16]. Evidence suggests that human papillomavirus vaccination and screening can prevent up to 95% of cervical cancer cases and reduce mortality by as much as 80% [17]. In Kerala, India, periodic mass and targeted oral visual screening in high-risk populations significantly reduced oral cancer mortality [18]. To scale up such preventive efforts, the Government of India introduced population-based screening under the National Health Mission through the National Programme for Prevention and Control of Noncommunicable Diseases (NP-NCD), implemented via Ayushman Bharat Health and Wellness Centres [19].
To assess the gap between policy intent and field-level implementation of population-based cancer screening, the National Family Health Survey-5 (NFHS-5) offers a valuable platform for evaluating coverage and effectiveness [20]. Given India’s substantial contribution to the global cancer burden, these initiatives have the potential to significantly reduce cancer-related morbidity and mortality nationwide. However, a paucity of nationally representative analyses disaggregated by cancer type and state-level variation remains a critical research gap. Although NFHS-5 provides macro-level estimates, evidence on micro-level sociodemographic determinants influencing screening uptake is limited.
To address this gap, a cross-sectional analytical study using NFHS-5 data was conducted. The analysis was restricted to women aged 15-49 years and aimed to assess state-specific disparities in breast, cervical, and oral cancer screening coverage, as well as to identify key determinants associated with cancer screening patterns.
Materials and methods
The NFHS-5, conducted between 2019 and 2021, introduced questions on cancer screening for the first time. This study aimed to assess the prevalence of screening for cervical, breast, and oral cancers among 225,565 women respondents aged 15-49 years. The analysis included women with complete information on cancer screening outcomes and relevant covariates; respondents with missing, incomplete, or “don’t know” responses for key outcome or explanatory variables were excluded. NFHS-5 employed a multistage sampling design, selecting households from both rural and urban areas across all 707 districts in India. Data were collected using computer-assisted personal interviewing with standardized questionnaires adopted from the NFHS-5 biomarker module, which is publicly available and open access, allowing efficient data collection and real-time data entry [21].
Chi-square tests were used to examine associations between outcome and predictor variables. Binary logistic regression models were employed to assess the relationship between selected independent variables and the likelihood of undergoing different cancer screening tests. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated to determine the strength and direction of associations. The dependent variable was a binary indicator representing whether a woman had ever undergone screening for cervical, breast, or oral cancer, based on the survey question, “Have you ever undergone a screening test for cervical, breast, or oral cancer?” Responses were coded as yes = 1 (screened) and no = 0 (not screened).
A domain-wise modeling approach was adopted, with three separate multivariable logistic regression models constructed to evaluate the influence of distinct predictor groups.
Model 1 (sociodemographic determinants) assessed screening patterns based on sociodemographic characteristics. The independent variables included were age ( \begin{document}x_1\end{document} ), place of residence ( \begin{document}x_2\end{document} ), education ( \begin{document}x_3\end{document} ), caste ( \begin{document}x_4\end{document} ), religion ( \begin{document}x_5\end{document} ), and wealth index ( \begin{document}x_6\end{document} ).
Model 2 (reproductive determinants) evaluated screening patterns based on reproductive factors, including birth order ( \begin{document}x_1\end{document} ), history of terminated pregnancy ( \begin{document}x_2\end{document} ), current breastfeeding status ( \begin{document}x_3\end{document} ), current marital status ( \begin{document}x_4\end{document} ), and current co-residence with husband/partner ( \begin{document}x_5\end{document} ).
Model 3 (ancillary determinants) examined screening patterns based on ancillary factors, including working status ( \begin{document}x_1\end{document} ), body mass index (BMI) ( \begin{document}x_2\end{document} ), husband/partner’s education ( \begin{document}x_3\end{document} ), contraceptive use and intention ( \begin{document}x_4\end{document} ), exposure to family planning messages on radio in the last few months ( \begin{document}x_5\end{document} ), exposure to family planning messages on television in the last few months ( \begin{document}x_6\end{document} ), and exposure to family planning messages in newspapers or magazines in the last few months ( \begin{document}x_7\end{document} ).
For all three models, the general form of the logistic regression equation was
[\ln\left(\frac{p}{1-p}\right) = \alpha_0 + \sum_{i=1}^{k} \alpha_i x_i]
where \begin{document}p\end{document} denotes the probability of having ever undergone cancer screening, \begin{document}\ln\left(\frac{p}{1-p}\right)\end{document} represents the log odds (logit) of the outcome, and \begin{document}\alpha_0, \alpha_1, \ldots, \alpha_k\end{document} are the logistic regression coefficients corresponding to the predictor variables included in each model.
Results
Table 1 presents the distribution of cancer screening uptake across states and union territories and at the national level, based on the absolute number (N) of women interviewed. At the national level, among the 225,565 women surveyed, only 1.2% reported having ever undergone cervical cancer screening. Breast cancer screening was reported by 0.5% of respondents, while 1,421 women (0.6%) reported screening for oral cancer. Overall, cancer screening coverage in India remains markedly low. Cervical cancer screening was the most commonly reported modality, followed by oral cancer screening, whereas breast cancer screening was the least utilized.
States with higher screening uptake
Tamil Nadu reported the highest screening coverage across all three cancers, with 7.5% for cervical cancer, 3.5% for breast cancer, and 0.9% for oral cancer screening. The Andaman and Nicobar Islands demonstrated an exceptionally high oral cancer screening rate of 10.26% (n = 4), substantially exceeding the national average. Maharashtra also showed relatively higher participation, with screening rates of 2.3% (n = 405) for cervical cancer, 1.5% (n = 276) for breast cancer, and 1.9% (n = 336) for oral cancer. Andhra Pradesh reported above-average uptake, particularly for cervical (3.0%) and oral cancer (3.1%) screening. Puducherry similarly demonstrated elevated screening levels, with 5.7% (n = 8) reporting cervical cancer screening and 2.9% (n = 4) reporting breast cancer screening.
States with lower screening uptake
Several states and UTs, including Dadra and Nagar Haveli and Daman and Diu, Ladakh, Lakshadweep, Assam, and Nagaland, reported either zero or negligible participation across all three cancer screening modalities. In West Bengal, among 15,542 women interviewed, only 14 (0.09%) reported cervical cancer screening, 15 (0.10%) reported breast cancer screening, and 12 (0.08%) reported oral cancer screening, indicating extremely low coverage.
Regional patterns
Distinct regional disparities in cancer screening uptake were observed. Northern and central Indian states, such as Uttar Pradesh, Rajasthan, Bihar, Delhi, Haryana, and Madhya Pradesh, exhibited substantial variability but generally low to moderate screening coverage. Despite their large populations, Rajasthan and Uttar Pradesh demonstrated particularly poor participation. In Uttar Pradesh, cervical cancer screening was reported by 0.9% of women, breast cancer screening by 0.3%, and oral cancer screening by 0.5%.
In contrast, southern and western states, including Tamil Nadu, Kerala, Maharashtra, Andhra Pradesh, and Goa, demonstrated relatively higher screening uptake, with Tamil Nadu, Kerala, and Maharashtra showing the most consistent coverage across cancer types. Among the northeastern states, Mizoram (4.0%), Arunachal Pradesh (0.6%), and Manipur (1.1%) reported comparatively higher cervical cancer screening rates. However, Nagaland and Sikkim reported no participation in breast or oral cancer screening (0%), highlighting potential gaps in access to and utilization of preventive services. Overall, cancer screening coverage was highest in the southern and western regions of India, while the northern, central, and northeastern regions continued to lag behind.
Table 2 presents the sociodemographic, reproductive, and ancillary correlates of ever undergoing cervical, breast, and oral cancer screening. The majority of variables demonstrated statistically significant associations with all three screening outcomes, as determined by the chi-square test (p < 0.05).
Table 2: Associates of women who have ever undergone screening tests for cervical, breast, and oral cancersχ²: chi-square test used to assess associations between predictor variables and cancer screening status, p < 0.05: statistically significant, n: absolute number of women who reported ever undergoing cervical cancer screening, %: corresponding proportion within each category
Screening uptake was significantly higher among urban women compared with rural counterparts (cervical: 1.6% vs. 1.0%; breast: 0.9% vs. 0.4%; oral: 0.9% vs. 0.5%; p = 0.00-0.032). Women aged 35-49 years reported the highest cervical cancer screening (1.5%), whereas women under 20 years had the lowest uptake across all cancer types. A clear educational gradient was observed: screening was highest among women with higher education (cervical: 2.2%; breast: 1.2%; oral: 0.9%) and lowest among women with no formal education (≤0.5% across screenings). Socio-cultural disparities were also evident: cervical cancer screening ranged from 0.6% among scheduled tribe women to 1.4% among other backward classes (p = 0.00), and from 0.6% among Muslim women to 4.7% among Jain women. A pronounced wealth gradient was observed (p = 0.00), with the richest women reporting higher uptake (cervical: 1.9%; breast: 1.0%; oral: 1.0%) than the poorest women (≤0.6% across screenings).
Regarding reproductive factors, screening was significantly higher among women with first-order births (breast: 0.6%; oral: 0.7%; p < 0.01), among women not currently breastfeeding (cervical: 1.5% vs. 1.0%; p = 0.00), and among women using or intending to use modern contraceptive methods (p = 0.00). A history of terminated pregnancy was associated with slightly higher cervical cancer screening (1.3%, n = 457; p = 0.03). Married women and those co-residing with their husband or partner also demonstrated higher screening uptake (p < 0.05).
Among ancillary factors, employment status was not significantly associated with screening (p > 0.05). In contrast, BMI showed a positive association, with cervical cancer screening highest among obese women (2.4%). The partner’s education was also influential, with cervical screening rates reaching 2.1% among women whose partners had higher education (p < 0.05).
Table 3 presents the multivariable logistic regression results examining predictors of cervical, breast, and oral cancer screening across three sequential models. Several sociodemographic, reproductive, and ancillary factors remained independently associated with screening uptake after adjustment.
Model 1 (sociodemographic determinants)
Rural residence was consistently associated with significantly lower odds of screening for all three cancers compared with urban residence (OR range: 0.68-0.87). Women aged 35-49 years had higher odds of cervical and breast cancer screening than those aged <20 years, whereas age was not significantly associated with oral cancer screening. Educational attainment demonstrated a strong positive association with screening uptake, with women having higher education exhibiting substantially increased odds across all three cancer screenings (OR range: 1.49-3.48).
Caste and religion remained significant predictors of screening behavior. Scheduled Tribe women and Muslim women had lower odds of undergoing cancer screening, while Sikh and Christian women showed higher odds of cervical cancer screening. Wealth status was independently associated only with cervical cancer screening, with women in higher wealth quintiles demonstrating increased odds of screening.
Model 2 (reproductive determinants)
Higher birth order was inversely associated with cancer screening uptake. For cervical cancer, the odds of screening declined progressively from second- or third-births (OR: 0.88; 95% CI: 0.80-0.95) to six or more births (OR: 0.55; 95% CI: 0.40-0.76), with similar inverse trends observed for breast and oral cancer screening. Currently, breastfeeding women have significantly lower odds of screening for cervical (OR: 0.63; 95% CI: 0.59-0.69), breast (OR: 0.59; 95% CI: 0.53-0.66), and oral cancer (OR: 0.84; 95% CI: 0.76-0.94).
Model 3 (ancillary factors)
Employment status was not significantly associated with cancer screening uptake. In contrast, BMI was positively associated with screening behavior. Overweight women had higher odds of undergoing cervical (OR: 1.78), breast (OR: 3.24), and oral cancer screening (OR: 1.86), while obesity was strongly associated with breast cancer screening (OR: 2.74).
Compared with modern contraceptive users, women using traditional contraceptive methods had significantly lower odds of cervical cancer screening (OR: 0.59; p < 0.001), with no significant differences observed for breast or oral cancer screening. Women not intending to use contraception had reduced odds of cervical (OR: 0.76; p = 0.01) and breast cancer screening (OR: 0.67; p = 0.03). Conversely, women intending to use contraception in the future showed higher odds of oral cancer screening (OR: 1.78; p = 0.02).
Exposure to family planning messages via radio was associated with higher odds of breast (OR: 2.35; p < 0.001) and oral cancer screening (OR: 2.47; p < 0.001), while the association with cervical cancer screening was borderline significant (OR: 1.26; p = 0.06). Television exposure was consistently associated with increased odds of screening across all three cancer types, including cervical (OR: 1.63; p < 0.001), breast (OR: 2.09; p < 0.001), and oral cancer screening (OR: 1.66; p = 0.01). In contrast, exposure to family planning messages through print media (newspapers or magazines) was not significantly associated with any screening outcome.
The partner’s educational attainment demonstrated a strong, graded association with screening behavior. Compared with women whose husbands or partners had no formal education, higher partner education was significantly associated with increased odds of cervical cancer screening, rising from primary (OR: 2.06; 95% CI: 1.35-3.14; p < 0.001) to secondary (OR: 1.53; 95% CI: 1.06-2.22; p = 0.02) and higher education (OR: 2.50; 95% CI: 1.68-3.74; p < 0.001). For breast cancer screening, significant associations were observed among women whose partners had secondary (OR: 2.31; 95% CI: 1.18-4.51; p = 0.01) and higher education (OR: 3.61; 95% CI: 1.78-7.31; p < 0.001), whereas primary education was not significant. In contrast, oral cancer screening showed no significant association with partner’s education across any category (p = 0.788).
Overall, Model 3 demonstrated the strongest and most consistent associations, underscoring the critical role of health communication exposure and household educational context in shaping cancer screening behavior.
Discussion
This study reveals perilously low national screening coverage for cervical (approximately three in 250 women), breast (one in 200), and oral (three in 500) cancers. These findings corroborate the findings of Borkotoky et al. (2024) and Gopika et al. (2022), both of whom documented widespread underutilization of early detection services in India [22,23]. Despite India’s high cancer burden and the availability of low-cost screening methods, particularly for oral and breast cancers, screening uptake remains alarmingly low. The extremely poor coverage of breast cancer screening underscores critical gaps in early detection efforts under the NP-NCD. Limited programmatic emphasis, inadequate integration within primary healthcare, and low population-level awareness likely contribute to these gaps. Moreover, disruptions to routine preventive health services during the COVID-19 pandemic, which coincided with NFHS-5 fieldwork, may have further suppressed screening uptake.
Marked regional disparities were also evident. Consistent with the spatial clustering patterns reported by Monica and Mishra (2020), southern and western states, particularly Tamil Nadu and Maharashtra, demonstrated relatively higher screening uptake, while screening in northeastern, northern, and central regions remained largely imperceptible [24]. Moran’s I spatial analysis by Nilima et al. (2022) similarly identified “hot spots” in Tamil Nadu and Maharashtra and “cold spots” in eastern and northeastern states [25]. Kalaiarasi Muthu et al. (2024) reported breast cancer screening coverage of approximately 54% in southern zones compared with only 3-5% in central and eastern India, further validating this regional divide. Chakravarti and Patel emphasized the need for context-specific interventions by highlighting substantial district-level variation in Uttar Pradesh [26,27]. The relatively high oral cancer screening observed in the Andaman and Nicobar Islands may reflect localized public health initiatives, enhanced outreach through primary healthcare services, or improved access in smaller population settings; however, this finding should be interpreted cautiously, given the small sample size.
Among northeastern states, Mizoram emerged as an outlier, with comparatively higher screening uptake. Sen et al. (2022) attributed this to strong community engagement and the advantages of a smaller population, which enhance program visibility and reach [28]. In contrast, Changkun et al. (2022), using NFHS-4 data, reported substantially higher screening estimates (approximately 21% for cervical and 9% for breast cancer), which likely reflect broader and less specific definitions of “screening” compared with the more explicit and categorical questions used in NFHS-5 [29].
Pronounced sociodemographic inequities were evident in screening uptake. Urban residence, higher educational attainment, and greater wealth were consistently associated with increased screening, mirroring findings by Ooi et al. (2024), who reported education (OR: 1.88) and high income (OR: 2.27) as key predictors [30]. Similar urban-rural differentials have been documented in Nepal by Lamichhane et al. (2023) [31]. A systematic review by Islam et al. (2021) further identified limited awareness and low educational attainment as principal barriers to breast and cervical cancer screening across low- and middle-income countries [32].
The age-specific pattern observed in this study, peak cervical cancer screening among women aged 30-49 years, is consistent with findings reported by Yadav et al. (2024) [33]. Lower screening participation among Muslim women and Scheduled Tribes aligns with evidence from decomposition analyses by Nilima et al. (2022) and Garg et al. (2023), underscoring the influence of religion and caste on healthcare utilization [25,34]. Nagar et al. (2023) further highlighted the role of cultural beliefs and stigma as major deterrents to screening participation. Although some studies have reported positive associations between screening uptake and women’s employment status [35,36], the absence of a significant association in our analysis may reflect the fact that employment does not uniformly translate into financial autonomy or decision-making power over healthcare in many Indian households. Formal education and cancer awareness, however, remain robust facilitators of screening uptake, consistent with findings by Sawhney et al. (2023) [37].
Our findings also indicate that exposure to family planning messages via television and radio, used here as proxy indicators of broader health awareness, significantly increases the likelihood of screening uptake. While Srinath et al.’s (2023) “7A” accessibility framework does not explicitly emphasize mass media influences [38], the wider reach and accessibility of television and radio may explain their stronger associations than those of print media, which relies on literacy and has comparatively limited penetration. The positive association between a partner’s education and screening uptake highlights the critical role of household decision-making dynamics, supportive environments, and facilitated access to preventive healthcare. Similarly, associations with contraceptive use and higher BMI may reflect increased contact with healthcare services, thereby enhancing opportunities for screening.
Conversely, lower screening uptake among women with higher parity and those currently breastfeeding suggests that caregiving responsibilities and competing domestic demands constrain access to preventive health services. Consistent with Subramanian et al. (2024), no significant association was observed between screening uptake and either employment status or history of pregnancy termination [39].
This study uniquely mapped cancer screening patterns across all Indian states and key population subgroups using NFHS-5 data, a large, nationally representative sample. Nonetheless, several limitations warrant consideration. The analysis relied on self-reported data, which introduced the potential for misclassification, recall bias, and respondent misunderstanding, particularly because a single survey item captured multiple screening types. Additionally, NFHS-5 includes only women aged 15-49 years, whereas the risk of cervical and breast cancer extends beyond this age range. The cross-sectional design precludes causal inference, and generalizability beyond India is limited due to contextual differences in healthcare systems and sociocultural norms.
Taken together, although “true” screening levels may be underestimated, the findings unequivocally demonstrate extremely low and uneven cancer screening coverage across India. These results underscore the urgent need for strengthened, context-sensitive screening strategies to reduce preventable cancer morbidity and mortality nationwide.
Conclusions
Cancer screening patterns in India, as revealed by the model-based findings, are shaped by a complex interplay of sociodemographic, reproductive, and ancillary factors. Urban residence, higher educational attainment, and belonging to non-scheduled caste/scheduled tribe groups consistently emerged as the strongest predictors of higher screening uptake. Women aged 35-49 years and those in higher wealth quintiles were also more likely to undergo cervical and breast cancer screening. Notably, Muslim women were significantly less likely to participate in any form of cancer screening, underscoring persistent religious and cultural disparities in access to preventive healthcare services.
Reproductive factors, including higher birth order, current breastfeeding, and not co-residing with a husband or partner, were strong negative predictors of screening across all three cancer types. In contrast, a history of pregnancy termination was not significantly associated with screening behavior. Among ancillary factors, overweight status, exposure to family planning messages, particularly through radio and television, and higher educational attainment of the husband or partner were positively associated with cancer screening uptake. Conversely, use of traditional contraceptive methods or lack of intention to use contraception was associated with lower screening uptake, especially for cervical and breast cancers.
In light of the observed sociodemographic disparities, targeted interventions are essential to improve equity in cancer screening access. Priority strategies include deploying mobile screening units in underserved rural areas, strengthening routine screening integration within primary healthcare services, and sustaining community-based outreach through frontline health workers to engage women with lower educational attainment. Additionally, culturally tailored awareness initiatives addressing caste- and religion-specific barriers, along with strengthened community-level referral linkages, may further enhance screening uptake among disadvantaged populations.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Noncommunicable diseases 3 2025 World Health Organization(2024 2024 https://www.who.int/news-room/fact-sheets/detail/noncommunicable-diseases
- 2Data visualization tools for exploring the global cancer burden in 2022 3 2025 2022 https://gco.iarc.who.int/today/en
- 3Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries CA Cancer J Clin Bray F Laversanne M Sung H Ferlay J Siegel RL Soerjomataram I Jemal A 2292637420243857275110.3322/caac.21834 · doi ↗ · pubmed ↗
- 4The GLOBOCAN 2022 cancer estimates: data sources, methods, and a snapshot of the cancer burden worldwide Int J Cancer Filho AM Laversanne M Ferlay J 1336134615620253968849910.1002/ijc.35278 · doi ↗ · pubmed ↗
- 5Noncommunicable diseases country profiles 2018 3 2025 2018 https://www.who.int/publications/i/item/ncd-country-profiles-2018
- 6Noncommunicable diseases progress monitor 2025 7 2025 2020 https://www.who.int/publications/i/item/9789240105775
- 7Cancer incidence estimates for 2022 and projection for 2025: result from National Cancer Registry Programme, India Indian J Med Res Sathishkumar K Chaturvedi M Das P Stephen S Mathur P 59860715620223651088710.4103/ijmr.ijmr_1821_22PMC 10231735 · doi ↗ · pubmed ↗
- 8Financial toxicity in cancer care: prevalence, causes, consequences, and reduction strategies J Surg Oncol Lentz R Benson AB 3rd Kircher S 859212020193065018610.1002/jso.25374 · doi ↗ · pubmed ↗