Abstract
This cohort study evaluates 10 years of Medicare claims data for outpatient encounters related to actinic keratosis.
Actinic keratosis (AK) is common. However, basic epidemiologic data on AK, such as incidence rates, are limited.1,2 Herein, we used individual-level Medicare data to report AK incidences and treatment patterns.
Methods
For this cohort study, we used the Research Data Assistance Center to obtain a random sample of 100% Medicare claims data (2009-2018) of beneficiaries 65 years or older without Medicare Advantage coverage (fee-for-service only) but with continuous Parts A, B, and D coverage. The MD Anderson Cancer Center Institutional Review Board approved this study and waived the informed consent requirement because deidentified data were used. We followed the STROBE reporting guideline.
We identified outpatient encounters for AK diagnosis with International Classification of Diseases (ICD), Ninth Revision code 702.0 and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision code L57.0, which were validated in a single-center study using electronic health record data.3 Outpatient encounters included professional and facility claims. Using χ2 tests, we compared demographic characteristics of beneficiaries with vs without AK diagnosis. We evaluated clinical characteristics of beneficiaries, including dermatological visits and skin cancer history. We descriptively evaluated potential AK treatments. eTables 1 to 3 in the Supplement list ICD and other codes.
We calculated raw incidence rates (number of AK diagnosis encounters per 100 000 person-years) overall, by sex, and by race and ethnicity (as defined by Research Triangle Institute race codes). We calculated age-adjusted incidence rates using the 2000 US standard population.4 We calculated 2 sets of incidence rates: 1 using AK diagnosis encounters, and 1 using AK diagnosis encounters with potential treatment (destruction, photodynamic therapy, or topical medications) to increase the likelihood AK was present because the positive predictive value of AK ICD codes was 90.5%.3
Two-sided P < .05 was considered significant. Statistical analysis was conducted using SAS 7.15 (SAS Institute).
Results
Of 4 999 999 beneficiaries (mean [SD] age, 72.3 [8.5] years; 3 108 772 women [62.2%]) with a mean (SD) follow-up of 5.74 (3.55) years, 1 462 985 (29.3%) had 1 or more AK diagnosis encounters (Table 1). We identified 8 178 543 AK diagnosis encounters (mean [SD], 5.59 [6.83] per patient), most commonly with dermatologists (n = 6 179 887 [78.6%]) and physician assistants or nurse practitioners (n = 852 900 [10.4%]). Topical fluorouracil was used in 186 715 encounters (2.3%); Table 2 shows treatment details.
Table 1. Demographic and Clinical Characteristics.
| Medicare beneficiaries, No. (%) | P valuea | |||
|---|---|---|---|---|
| With at least 1 AK diagnosis | Without AK diagnosis | All beneficiaries | ||
| Total patients | 1 462 985 | 3 512 658 | 4 999 999 | |
| Sex | ||||
| Female | 806 813 (55.2) | 2 288 866 (65.2) | 3 108 772 (62.2) | <.001 |
| Male | 656 172 (44.9) | 1 223 792 (34.8) | 1 891 227 (37.8) | |
| Race and ethnicityb | ||||
| American Indian/Alaska Native | 2278 (0.2) | 19 786 (0.6) | 22 166 (0.4) | <.001 |
| Asian/Pacific Islander | 9644 (0.7) | 112 588 (3.2) | 122 365 (2.5) | |
| Black or African American | 3186 (0.2) | 345 162 (9.8) | 348 697 (7.0) | |
| Hispanic | 23 908 (1.6) | 219 963 (6.3) | 244 741 (4.9) | |
| Non-Hispanic White | 1 398 890 (95.6) | 2 749 922 (78.3) | 4 171 180 (83.4) | |
| Otherc | 5194 (0.4) | 23 849 (0.7) | 29 261 (0.6) | |
| Unknown | 19 885 (1.4) | 41 388 (1.2) | 61 589 (1.2) | |
| Census region | ||||
| Midwest | 330 267 (22.6) | 924 868 (26.3) | 1 260 259 (25.2) | <.001 |
| South | 612 313 (41.9) | 1 313 048 (37.4) | 1 935 683 (38.7) | |
| Northeast | 229 926 (15.7) | 692 488 (19.7) | 926 328 (18.5) | |
| West | 289 255 (19.8) | 576 048 (16.4) | 870 167 (17.4) | |
| Other or unknownd | 1224 (0.1) | 6206 (0.2) | 7562 (0.2) | |
| Age at first AK diagnosis, y | ||||
| 65-69 | 535 722 (36.6) | NA | NA | NA |
| 70-74 | 293 159 (20.0) | |||
| 75-79 | 249 001 (17.0) | |||
| 80-84 | 196 446 (13.4) | |||
| 85-89 | 123 131 (8.4) | |||
| ≥90 | 65 526 (4.5) | |||
| Follow-up after first AK, mean (SD), y | 5.23 (3.1) | NA | NA | NA |
| Saw a dermatologist at any timee | ||||
| No | 293 863 (20.1) | NA | NA | NA |
| Yes | 1 169 122 (79.9) | |||
| History of skin cancerf,g | ||||
| No | 1 075 046 (73.5) | NA | NA | NA |
| Yes | 387 939 (26.5) | |||
| Keratinocyte carcinomag | 125 622 (8.6) | NA | NA | NA |
| Basal cell carcinoma | 61 951 (4.2) | NA | NA | NA |
| Squamous cell carcinoma | 66 429 (4.5) | NA | NA | NA |
| Melanoma | 46 144 (3.2) | NA | NA | NA |
| Unknown type | 241 778 (16.5) | NA | NA | NA |
Abbreviations: AK, actinic keratosis; NA, not applicable.
χ2 was used to compare beneficiaries with at least 1 AK diagnosis with beneficiaries without AK diagnosis. Beneficiaries with at least 1 AK diagnosis had at least 1 AK diagnosis in an outpatient encounter when the patient was 65 years or older (Medicare eligible based on age). Beneficiaries without AK diagnosis had no AK diagnoses in any encounter setting (including inpatient) or age (for some patients, data included claims information at age <65 y).
Race and ethnicity were defined by the Research Triangle Institute (RTI) race code in the Medicare Master Beneficiary Summary File.
All race and ethnicity categories, including other, reflect the terms used by RTI.
Other or unknown includes any missing value, Guam, Puerto Rico, and US Virgin Islands.
Dermatologists identified by the National Provider Index taxonomy code.
Skin cancer history was defined by an International Classification of Diseases, Ninth Revision or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision code for skin cancer before the initial AK diagnosis or for history of skin cancer before or on the same day as the initial AK diagnosis (eTable 2 in the Supplement).
Number of patients with keratinocyte carcinoma, basal cell carcinoma, squamous cell carcinoma, melanoma, and unknown type was not exclusive and added up to more than the total number of patients with a history of skin cancer.
Table 2. Characteristics of Encounters With AK Diagnoses.
| No. (%) | ||||
|---|---|---|---|---|
| All encounters | Encounters in patients without history of skin cancer | Encounters in patients with history of skin cancer | Encounters in patients with history of SCC | |
| Total encounters | 8 178 543 | 5 838 872 | 2 339 671 | 288 811 |
| With potential treatmenta | 6 697 341 (81.9) | 4 767 085 (81.6) | 1 930 256 (82.5) | 221 551 (76.7) |
| Destructionb | ||||
| 1 AK treated | 1 545 313 (18.9) | 1 093 674 (18.7) | 451 639 (19.3) | 58 596 (20.3) |
| 2-14 AKs treated | 4 085 718 (50.0) | 2 905 157 (49.8) | 1 180 561 (50.5) | 128 619 (44.5) |
| ≥15 AKs treated | 825 819 (10.1) | 600 785 (10.3) | 225 034 (9.6) | 23 837 (8.3) |
| Photodynamic therapyc | 121 572 (1.5) | 83 374 (1.4) | 38 198 (1.6) | 5021 (1.7) |
| Topical medicationd | 240 122 (2.9) | 166 591 (2.9) | 73 531 (3.1) | 10 389 (3.6) |
| Fluorouracil | 186 715 (2.3) | 129 915 (2.2) | 56 800 (2.4) | 8110 (2.8) |
| Imiquimod | 41 500 (0.5) | 27 754 (0.5) | 13 746 (0.6) | 1931 (0.7) |
| Ingenol mebutate | 5783 (0.1) | 4055 (0.1) | 1728 (0.1) | 232 (0.1) |
| Diclofenac, 3% | 7154 (0.1) | 5536 (0.1) | 1618 (0.1) | 180 (0.1) |
| Without potential treatment | 1 481 202 (18.1) | 1 071 787 (18.4) | 409 415 (17.5) | 67 260 (23.3) |
Abbreviations: AK, actinic keratosis; SCC, squamous cell carcinoma.
Percentages add up to more than 100% because these treatment categories are not mutually exclusive (eg, an encounter could have a destruction code and also a topical medication prescription).
1 AK treated (Current Procedural Terminology code 17000 alone); 2-14 AKs treated (Current Procedural Terminology code 17003); ≥15 AKs treated (Current Procedural Terminology code 17004 alone).
Photodynamic therapy was defined by Current Procedural Terminology code 96567, 96570, 96571, 96573, 96574, J7308, or J7309.
Topical medications were identified by National Drug Code codes present on the day of the encounter with an AK diagnosis or in the 2 weeks after the encounter (eTable 3 in the Supplement).
Raw incidence rate of AK diagnosis encounters was 28 656 per 100 000 person-years (42 970 for men, 20 492 for women; 32 664 for non-Hispanic White patients, 4338 for all other race and ethnicity groups; 28 788 age-adjusted rate). Raw incidence rate of AK diagnosis encounters with treatments was 23 466 per 100 000 person-years (36 147 for men, 16 234 for women; 26 786 for non-Hispanic White patients, 3305 for all other race and ethnicity groups; 23 625 age-adjusted rate).
Discussion
Three in 10 beneficiaries had at least 1 AK diagnosis encounter over 5 years of follow-up. Incidence rates were high: 2 in 5 men and 1 in 5 women had AK diagnosis encounters per year. These rates likely underestimated the number of individual AK lesions.
Topical field treatments had low utilization (<5% of encounters): fluorouracil, recently associated with decreased squamous cell carcinoma incidence,5 was used in less than 3% of encounters. This finding might reflect patient preferences6 or financial incentives with cryotherapy. No meaningful treatment differences by skin cancer history were observed.
This study provides AK incidence rates that were not previously available but has limitations. There was potential for inaccurate ICD coding (ie, AK ICD codes without AK), which we addressed by evaluating encounters with potential treatments (>80% of encounters). We could not exclude patients with AK history. We included only fee-for-service coverage, which limits generalizability.
Actinic keratosis is a major public health problem for US older adults. The findings demonstrated high incidence rates and suggested an opportunity for squamous cell carcinoma prevention through increased use of topical fluorouracil.
eTable 1. Skin Cancer Diagnosis Codes for Definition of History of Skin Cancer
eTable 2. History of Skin Cancer ICD-9 and ICD-10 Codes
eTable 3. NDC Code List
References
- 1.Gupta AK, Cooper EA, Feldman SR, Fleischer AB Jr. A survey of office visits for actinic keratosis as reported by NAMCS, 1990-1999. National Ambulatory Medical Care Survey. Cutis. 2002;70(2 suppl):8-13. [PubMed] [Google Scholar]
- 2.Yeung H, Baranowski ML, Swerlick RA, et al. Use and cost of actinic keratosis destruction in the Medicare Part B fee-for-service population, 2007 to 2015. JAMA Dermatol. 2018;154(11):1281-1285. doi: 10.1001/jamadermatol.2018.3086 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cohen OG, Margolis DJ, Wehner MR. The validity of diagnostic and treatment codes for actinic keratosis in electronic health records. Br J Dermatol. 2020;182(6):1487-1488. doi: 10.1111/bjd.18779 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.National Cancer Institute. Standard populations (millions) for age-adjustment. Accessed February 11, 2022. https://seer.cancer.gov/stdpopulations/
- 5.Weinstock MA, Thwin SS, Siegel JA, et al. ; Veterans Affairs Keratinocyte Carcinoma Chemoprevention Trial (VAKCC) Group . Chemoprevention of basal and squamous cell carcinoma with a single course of fluorouracil, 5%, cream: a randomized clinical trial. JAMA Dermatol. 2018;154(2):167-174. doi: 10.1001/jamadermatol.2017.3631 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kopasker D, Kwiatkowski A, Matin RN, et al. Patient preferences for topical treatment of actinic keratoses: a discrete-choice experiment. Br J Dermatol. 2019;180(4):902-909. doi: 10.1111/bjd.16801 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Skin Cancer Diagnosis Codes for Definition of History of Skin Cancer
eTable 2. History of Skin Cancer ICD-9 and ICD-10 Codes
eTable 3. NDC Code List
