Abstract
Introduction:
People with HIV (PWH) are more likely to die due to cancer compared to people without HIV. Disparities in cancer care, including access to palliative care, for PWH contribute to poor health-related quality of life (HRQoL) and survival. However, limited research exists examining patient-reported HRQoL among PWH with cancer. We examined HRQoL among patients diagnosed with non-AIDS defining cancers with and without HIV.
Materials and methods:
We used the 2007–2017 Surveillance, Epidemiology, and End Results and the Medicare Health Outcomes Survey (SEER-MHOS) linkage data to assess HRQoL among patients (ages ≥18) diagnosed with breast, colorectal, gastrointestinal, head and neck, lung, lymphoma, and prostate cancers. HRQoL outcomes included the Physical Component Summary (PCS) and Mental Component Summary (MCS) scores and the eight scale scores of the Veterans RAND 12-Item Health Survey (VR-12), and the VR-6D health utility score. Higher HRQoL scores indicate better health status. Adjusting for patient characteristics, we computed mean HRQoL scores using multivariable linear regression models and the predictive margins method. The minimally important difference (MID) in HRQoL scores between patients with and without HIV by cancer types was assessed.
Results:
The sample (N = 43,973) included 310 (0.7%) patients with HIV and had an average age at cancer diagnosis of 70.8 years. PWH reported lower scores in at least one HRQoL outcome compared to patients without HIV for all cancers examined. Differences in HRQoL for PWH compared to patients without HIV exceeded the MID for cancers of the breast (PCS: −2.7; MCS: −9.1; VR-6D: −0.08), colorectum (PCS: −2.3; MCS: −6.2; VR-6D: −0.06), gastrointestinal tract (MCS: −5.9; VR-6D: −0.04), head and neck (PCS: −4.1; MCS: −7.5; VR-6D: −0.07), lungs (PCS: −2.5; MCS: −6.3; VR-6D: −0.06), lymphatic system (PCS: −2.3; MCS: −2.6; VR-6D: −0.04), and prostate (MCS: −8.0; VR-6D: −0.07).
Discussion:
Our findings demonstrated that PWH across all cancers examined reported substantially lower mental and/or physical HRQoL compared to patients without HIV. Future work can explore strategies for symptom monitoring and management among PWH.
Keywords: HIV, Cancer, Health-related quality of life, SEER-MHOS, Disparities
1. Introduction
People with HIV (PWH) face disproportionately higher rates of certain cancers and cancer mortality compared to people without HIV [1,2]. Antiretroviral therapies (ARTs) have improved the quality of life and overall survival among PWH [3]. As the aging population of PWH grows, cancer has become the leading cause of non-AIDS death among PWH in the United States (US) [1,4,5]. Historically, the most common cancers among PWH were AIDS-defining cancers (e.g., Kaposi sarcoma) that have an infectious etiology traditionally linked to living with HIV [1,4,5]. However, the leading causes of cancer deaths among PWH in the US now are largely cancers without an infectious etiology (e.g., lung and prostate cancers), collectively known as non-AIDS defining cancers [1,4,5]. Elevated mortality rates among PWH with cancer have multifaceted causes and are attributable, in part, to inequities in cancer care delivery [6,7]. In fact, PWH are less likely to receive supportive care during cancer treatment, which may contribute to the poorer cancer outcomes observed among this understudied cancer patient population [8–10].
Patient-reported health-related quality of life (HRQoL), including physical and psychosocial functioning, are important measures in evaluating impacts of cancer care on a patient’s well-being and can serve as prognostic factors for a cancer patient’s risk of morbidity and mortality [11,12]. Poor symptom management and lower HRQoL are highly associated with increased preventable hospitalizations, treatment discontinuation, and mortality [13–15]. Patients with cancer across multiple cancer types exhibit lower HRQoL compared to the general population [12,16,17]. However, limited research exists examining patient-reported HRQoL among PWH with cancer. There is an urgent need to evaluate the HRQoL among PWH with cancer given the increased risk for reduced HRQoL due to comorbid HIV and cancer. Improving HRQoL may be an intervenable opportunity to improve cancer outcomes among PWH. We aimed to assess differences in HRQoL among a national sample of adults with non-AIDS-defining cancers with and without HIV.
2. Methods
2.1. Data
We analyzed linkage data between the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) national cancer registry and the Centers for Medicare and Medicaid Services’ Medicare Health Outcomes Survey (MHOS) [18]. The SEER-MHOS dataset contains sociodemographic, health characteristics, and patient-reported health outcomes of randomly sampled adults enrolled in Medicare Advantage plans [18]. Although MHOS respondents are Medicare-eligible individuals who are primarily age 65 and older, those younger than 65 may be included based on disability status or specific clinical conditions, such as advanced HIV or AIDS [19]. States and metropolitan areas participating in SEER included New York (24% of participants), California (16%), Georgia (8%), Massachusetts (6%), Louisiana (5%), New Jersey (5%), Seattle-Puget Sound (5%), Connecticut (4%), Hawaii (4%), Idaho (4%), New Mexico (4%), Utah (4%), Detroit (3%), Iowa (3%), and Kentucky (3%), with 2% unknown. The response rate for the MHOS across the study years averaged 52.9% (standard deviation [SD]: 7.8; range: 43.0%–63.4%) [20]. This study was reviewed and deemed exempt from formal review by the Moffitt Cancer Center Institutional Review Board of Record, Advarra.
2.2. Study cohort
We included 11 cohorts of participants between January 2007 and December 2017 of participants ages 18 and older diagnosed with seven cancers common among PWH identified using the SEER cancer site recodes that were mapped to the International Classification of Diseases for Oncology, 3rd Edition and the World Health Organization 2008 site and histology codes (Appendix A) [21]. Cancers examined included breast, colorectal, gastrointestinal, head and neck, lung, lymphoma, and prostate. For each participant, we included data from the MHOS survey completed in the same month as their cancer diagnosis or the first survey completed following their cancer diagnosis. Participants of the MHOS survey provided informed consent. The SEER-MHOS linkage data are considered a limited dataset with no personal identifying information and is therefore exempt from additional informed consent requirements.
2.3. Measures
2.3.1. HIV status
Similar to the approach used in previous literature, we determined a participant’s HIV-positive status based on their use of HIV-specific antiretroviral therapy (ART) medications [22]. Specifically, participants who have used HIV-specific medications listed in Appendix B were classified as having HIV. Exceptions were made for participants who used ART medications for the treatment of hepatitis B or for pre-exposure prophylaxis (PrEP) to prevent HIV (lamivudine alone; emtricitabine alone; tenofovir disoproxil fumarate alone or in combination with emtricitabine; and tenofovir alafenamide alone or in combination with emtricitabine) [22]. We also assessed participants for the use of post-exposure prophylaxis (PEP) to prevent HIV after possible exposure (defined as a single 28-day regiment of a combination of ART medications) to determine the potential for misclassifying HIV-negative participants taking PEP as being HIV positive [23]. No participants using PEP were identified in our study.
2.3.2. Outcomes
The primary outcomes in our study included the eight scale of the Veterans RAND 12-Item Health Survey (VR-12) assessing participants’ HRQoL, which include the following domains: physical functioning, role limitations due to physical health problems (hereafter, “role physical”), bodily pain, general health, vitality, social functioning, role limitations due to personal or emotional problems (hereafter, “role emotional”), and mental health [24]. The Physical Component Summary (PCS) and Mental Component Summary (MCS) scores were also computed from the VR-12 scale scores [25]. The scale scores and the summary scores were standardized to the general US population norms using a T score metric (range 0–100, with mean score of 50 and standard deviation of 10). Higher scores indicate better HRQoL. Consistent with previous studies [26], the minimally important difference (MID), representing the minimal change in a participant’s health status to be considered meaningful, was defined as 3 points or greater in the eight scale scores and 2 points or greater in the PCS and MCS scores. We also computed the Veterans RAND 6-Dimension Health Survey (VR-6D), a single preference-based health utility measure with scores ranging from 0 (death) to 1 (perfect health) derived from the VR-12 [27,28]. The MID for the VR-6D score was defined as 0.03 or greater [29,30].
2.4. Statistical analyses
Participant characteristics were described overall and by HIV status. Participant counts smaller than 11 were suppressed to protect the identities of participants. Adjusted mean PCS and MCS scores, the eight scale scores, and the VR-6D scores with 95% confidence interval (95% CI) were computed using multivariable linear regression models and the predictive margins method with marginal standardization, averaging predicted outcomes over the observed covariate distribution within the study sample [31–33]. Although the differences in adjusted mean scores were presented with 95% CIs, our study interpreted meaningful differences between participants with and without HIV based on the MID established in prior literature as described above [26,29,30]. The MID is widely accepted in its use for studies examining patient-reported outcomes due to its emphasis on clinical relevance and patient experience without the drawback of dependency on sample size as with statistical significance [34,35]. All models were adjusted for age at first cancer diagnosis, sex, race/ethnicity, highest educational attainment, marital status, annual household income, cancer stage at diagnosis, whether the participant had multiple cancer diagnoses, whether the participant was diagnosed with any of 12 chronic conditions (hypertension, arthritis of the hip or knee, arthritis of the hand or wrist, diabetes, sciatica, chronic obstructive pulmonary disease, coronary artery disease, myocardial infarction, congestive heart failure, stroke, other heart conditions, and inflammatory bowel disease), months from cancer diagnosis to survey completion, mode of survey administration, and whether the survey was completed by a proxy respondent. A complete case approach was used for analysis, with missing values at less than 10% and summarized in Table 1. All analyses were performed in RStudio (Posit, Boston, MA, USA) and SAS v.9.4 (SAS Institute, Cary, NC, USA). Figures were generated using GraphPad Prism v.10 (GraphPad Software, Boston, MA, USA).
Table 1.
Characteristics of participants overall and by HIV status.
| Characteristic | No. (%) or Mean (SD) | ||
|---|---|---|---|
| Overall (N = 43,973) | HIV negative (N = 43,663) | HIV positive (N = 310) | |
| Sociodemographic | |||
| Age at cancer diagnosis, years | |||
| 18–65 | 8464 (19.2) | 8243 (18.9) | 221 (71.3) |
| ≥65 | 35,509 (80.8) | 35,420 (81.1) | 89 (28.7) |
| Mean (SD) | 70.8 (8.58) | 70.9 (8.49) | 57.6 (10.4) |
| Sex | |||
| Male | 22,542 (51.3%) | 22,295 (51.1%) | 247 (79.7%) |
| Female | 21,431 (48.7%) | 21,368 (48.9%) | 63 (20.3%) |
| Race/ethnicity | |||
| White, non-Hispanic | 32,274 (73.4%) | 32,164 (73.7%) | 110 (35.5%) |
| Black or other, non-Hispanic | 10,206 (23.2%) | 10,041 (23.0%) | 165 (53.2%) |
| Hispanic/Latinx | 1493 (3.40%) | 1458 (3.34%) | 35 (11.3%) |
| Highest educational attainment | |||
| Some high school or less or unknown | 11,812 (26.9%) | 11,720 (26.8%) | 92 (29.7%) |
| High school or equivalent | 13,284 (30.2%) | 13,209 (30.3%) | 75 (24.2%) |
| Some college or 2-year degree | 10,266 (23.3%) | 10,186 (23.3%) | 80 (25.8%) |
| 4-year college or more | 8611 (19.6%) | 8548 (19.6%) | 63 (20.3%) |
| Marital status | |||
| Never married | 2929 (6.66%) | 2781 (6.37%) | 148 (47.7%) |
| Married | 22,290 (50.7%) | 22,250 (51.0%) | 40 (12.9%) |
| Separated | 1134 (2.58%) | 1105 (2.53%) | 29 (9.35%) |
| Divorced | 6624 (15.1%) | 6584 (15.1%) | 40 (12.9%) |
| Widowed | 9793 (22.3%) | 9757 (22.3%) | 36 (11.6%) |
| Unknown | 1203 (2.74%) | 1186 (2.72%) | 17 (5.48%) |
| Annual household income | |||
| Less than $5000 | 2402 (5.46%) | 2361 (5.41%) | 41 (13.2%) |
| $5000–$9999 | 3334 (7.58%) | 3272 (7.49%) | 62 (20.0%) |
| $10,000–$19,999 | 8019 (18.2%) | 7932 (18.2%) | 87 (28.1%) |
| $20,000 or more | 21,082 (47.9%) | 21,018 (48.1%) | 64 (20.6%) |
| Unknown | 9136 (20.8%) | 9080 (20.8%) | 56 (18.1%) |
| Urbanicity of residence | |||
| Big metropolitan area (counties with ≥1 million population) | 25,504 (58.0%) | 25,239 (57.8%) | 265 (85.5%) |
| Non-big metropolitan area or unknown | 18,469 (42.0%) | 18,424 (42.2%) | 45 (14.5%) |
| Home ownership | |||
| Not owned or rented | 10,545 (24.0%) | 10,342 (23.7%) | 203 (65.5%) |
| Owned | 28,846 (65.6%) | 28,786 (65.9%) | 60 (19.4%) |
| None of the above | 2207 (5.02%) | 2183 (5.00%) | 24 (7.74%) |
| Unknown | 2375 (5.40%) | 2352 (5.39%) | 23 (7.42%) |
| Health characteristics | |||
| Cancer site | |||
| Breast | 12,828 (29.2%) | 12,801 (29.3%) | 27 (8.71%) |
| Colorectal | 5968 (13.6%) | 5943 (13.6%) | 25 (8.06%) |
| Gastrointestinal | 2437 (5.54%) | 2338 (5.35%) | 99 (31.9%) |
| Head and neck | 1818 (4.13%) | 1799 (4.12%) | 19 (6.13%) |
| Lung | 4705 (10.7%) | 4683 (10.7%) | 22 (7.10%) |
| Lymphoma | 2597 (5.91%) | 2557 (5.86%) | 40 (12.9%) |
| Prostate | 13,620 (31.0%) | 13,542 (31.0%) | 78 (25.2%) |
| Cancer stage at diagnosis | |||
| In situ or unknown | 4980 (11.3%) | 4898 (11.2%) | 82 (26.5%) |
| Localized | 25,922 (58.9%) | 25,788 (59.1%) | 134 (43.2%) |
| Regional | 9073 (20.6%) | 9021 (20.7%) | 52 (16.8%) |
| Distant | 3998 (9.09%) | 3956 (9.06%) | 42 (13.5%) |
| Whether diagnosed with multiple cancers | |||
| No | 32,185 (73.2%) | 31,957 (73.2%) | 228 (73.5%) |
| Yes | 11,788 (26.8%) | 11,706 (26.8%) | 82 (26.5%) |
| Smoking status | |||
| No | 38,222 (86.9%) | 38,002 (87.0%) | 220 (71.0%) |
| Everyday or some days | 4647 (10.6%) | –a | –a |
| Unknown | 1104 (2.51%) | –a | –a |
| Diagnosed with any of the 12 chronic conditionsb | |||
| No | 4437 (10.1) | 4386 (10.0) | 51 (16.5) |
| Yes | 39,536 (89.9) | 39,277 (90.0) | 259 (83.5) |
| Survey characteristics | |||
| Time from cancer diagnosis to survey, months | 32.0 (28.1) | 31.9 (28.1) | 36.0 (28.8) |
| Mode of survey administration | |||
| 34,976 (79.5%) | 34,734 (79.6%) | 242 (78.1%) | |
| Telephone | 8997 (20.5%) | 8929 (20.4%) | 68 (21.9%) |
| Survey completed by a proxy respondent | |||
| No | 36,405 (82.8%) | 36,155 (82.8%) | 250 (80.6%) |
| Yes | 5377 (12.2%) | 5337 (12.2%) | 40 (12.9%) |
| Unknown | 2191 (4.98%) | 2171 (4.97%) | 20 (6.45%) |
Suppressed due to small cell.
Chronic conditions included hypertension, arthritis of the hip or knee, arthritis of the hand or wrist, diabetes, sciatica, chronic obstructive pulmonary disease, coronary artery disease, myocardial infarction, congestive heart failure, stroke, other heart conditions, and inflammatory bowel disease.
To compare participants with HIV in our sample to similar participants without HIV in order to reduce the potential effects of confounding [36–38], we performed a propensity score matching analysis (Appendix C). Using logistic regression models, we estimated propensity scores based on variables recommended by SEER-MHOS guidance [39], including age within five years, sex, race/ethnicity, cancer type, and year of cancer diagnosis. Using the nearest neighbor method without replacement, participants without HIV were matched 5:1 to participants with HIV, yielding a matched sample of 309 participants with HIV and 1528 participants without HIV (one participant with HIV was unable to be matched). Standardized mean differences across covariates were reduced after matching, with most covariates achieving an absolute standardized mean difference below 0.1. Covariate-specific density plots further confirmed improved balance across the matched samples.
3. Results
3.1. Sample characteristics
The study included 43,973 participants, and 310 (0.7%) participants were HIV positive (Table 1). The average age at cancer diagnosis was 70.8 years, and 80.8% of participants were age 65 and older. About half (51.3%) of participants were male, and the racial/ethnic composition of participants included non-Hispanic (NH)-White (73.4%), NH-Black or other race (23.2%), and Hispanic/Latinx (3.4%). The highest educational attainment of most participants was high school or equivalent (30.2%) or some college (23.3%). Half (50.7%) of participants were married, and 47.9% of participants had an annual household income of $20,000 or more.
Cancer sites of participants included prostate (31.0%), breast (29.2%), colorectal (13.6%), lung (10.7%), lymphoma (5.9%), gastrointestinal (5.5%), and head and neck (4.1%) (Table 1). Most participants responded to the survey via mail (79.5%) and completed the survey on their own (82.8%). The average time between cancer diagnosis to survey completion was 32.0 months. Select patient characteristics, including sex, age at cancer diagnosis, and time from cancer diagnosis to survey completion by cancer sites are shown in Table 2.
Table 2.
Select participant characteristics by cancer sites.
| Cancer site | Overall, No. | Female, % | Age at cancer diagnosis, years | Months from diagnosis to survey | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean (SD) | Median (IQR) | Mean (SD) | Median (IQR) | ||||||||||
| Total | HIV− | HIV+ | HIV− | HIV+ | HIV− | HIV+ | HIV− | HIV+ | HIV− | HIV+ | HIV− | HIV+ | |
| Total | 43,973 | 43,663 | 310 | 48.9 | 20.3 | 70.9 (8.5) | 57.6 (10.4) | 71 (66, 76) | 59 (51, 66) | 31.9 (28.1) | 36.0 (28.8) | 23 (10, 49) | 31 (11, 56) |
| Breast | 12,828 | 12,801 | 27 | 99.3 | 100.0 | 70.5 (9.0) | 58.7 (11.2) | 70 (65, 76) | 60 (50, 67) | 34.0 (28.4) | 40.7 (23.6) | 25 (11, 51) | 46 (24, 55.5) |
| Colorectal | 5968 | 5943 | 25 | 51.5 | 40.0 | 72.4 (9.2) | 59.8 (8.5) | 72 (67, 79) | 61 (55, 66) | 31.3 (27.6) | 29.0 (23.6) | 22 (10, 48) | 23 (11, 47) |
| Gastrointestinal | 2437 | 2338 | 99 | 44.8 | –a | 71.1 (9.2) | 55.1 (10.5) | 71 (66, 77) | 55 (50, 62) | 23.1 (24.5) | 36.1 (29.2) | 14 (5, 33) | 31 (11, 55) |
| Head and neck | 1818 | 1799 | 19 | 34.7 | –a | 69.5 (9.6) | 55.5 (10.7) | 70 (64, 76) | 55 (46, 67) | 31.3 (27.6) | 40.8 (23.0) | 22 (9, 48) | 41 (20, 58) |
| Lung | 4705 | 4683 | 22 | 55.5 | –a | 72.1 (8.2) | 60.5 (7.2) | 72 (67, 78) | 62 (57, 66.8) | 21.9 (23.1) | 27.6 (27.6) | 14 (5, 30) | 18 (8.5, 37) |
| Lymphoma | 2597 | 2557 | 40 | 51.7 | –a | 71.6 (9.7) | 51.2 (11.7) | 72 (66, 78) | 50.5 (44.5, 57) | 31.0 (27.8) | 51.5 (29.8) | 21 (9, 46) | 58 (24.8, 70.2) |
| Prostate | 13,620 | 13,542 | 78 | N/A | N/A | 70.3 (7.0) | 62.9 (7.2) | 70 (66, 75) | 63.5 (57, 67.8) | 35.6 (29.1) | 29.6 (29.9) | 27 (12, 54) | 20 (7, 41.5) |
Suppressed due to small cell.
3.2. Health-related quality of life
Table 3 summarizes the adjusted physical (PCS) and mental (MCS) symptom scores by cancer sites among patients with and without HIV. Among PWH, PCS scores were lowest for patients with lung cancer (30.5) (Table 3). PWH had lower MCS scores across all the cancers examined (range: 41.1–47.2) compared to patients without HIV (range: 47.3–51.0) (Table 3). Among PWH, MCS score was lowest for patients with breast (41.1) and head and neck (41.4) cancers. PWH had lower VR-6D scores across all cancers examined (range: 0.57–0.62) compared to patients without HIV (range: 0.63–0.69) (Table 3). Among PWH, VR-6D scores were lowest for patients with lung cancer (0.57). PWH with breast, colorectal, head and neck, lung, and lymphoma cancers had lower PCS, MCS, and VR-6D scores compared to patients without HIV (Table 3).
Table 3.
Adjusted mean Physical and Mental Component Summary scores and Veterans RAND 6-Dimension Health Survey (VR-6D) utility scores by HIV status and cancer sites.
| Cancer Site | HIV negative | HIV positive | Difference (95% CI) | ||
|---|---|---|---|---|---|
| N | Mean (95% CI) | N | Mean (95% CI) | ||
| Physical Component Summarya,b | |||||
| Breast | 12,793 | 37.3 (37.0, 37.5) | 27 | 34.6 (29.9, 39.4) | −2.7 (−7.4, 2.0) |
| Colorectal | 5938 | 37.1 (36.8, 37.4) | 25 | 34.8 (30.1, 39.4) | −2.3 (−7.0, 2.4) |
| Gastrointestinal | 2338 | 34.3 (33.9, 34.8) | 99 | 35.7 (33.4, 38.0) | 1.4 (−1.0, 3.8) |
| Head and neck | 1799 | 36.7 (36.1, 37.2) | 19 | 32.6 (28.7, 36.5) | −4.1 (−8.1, −0.1) |
| Lung | 4680 | 33.0 (32.6, 33.3) | 22 | 30.5 (26.4, 34.6) | −2.5 (−6.6, 1.6) |
| Lymphoma | 2556 | 36.3 (35.8, 36.7) | 40 | 34.0 (29.5, 38.4) | −2.3 (−6.8, 2.2) |
| Prostate | 13,533 | 38.5 (38.3, 38.8) | 78 | 37.5 (34.7, 40.3) | −1.0 (−3.8, 1.8) |
| Mental Component Summarya, b | |||||
| Breast | 12,790 | 50.2 (49.9, 50.4) | 27 | 41.1 (35.9, 46.2) | −9.1 (−14.2, −4.0) |
| Colorectal | 5939 | 50.0 (49.7, 50.3) | 25 | 43.8 (39.3, 48.2) | −6.2 (−10.7, −1.7) |
| Gastrointestinal | 2337 | 47.8 (47.3, 48.3) | 99 | 41.9 (39.2, 44.6) | −5.9 (−8.7, −3.1) |
| Head and neck | 1796 | 48.9 (48.4, 49.5) | 19 | 41.4 (35.7, 47.1) | −7.5 (−13.2, −1.8) |
| Lung | 4679 | 48.6 (48.2, 48.9) | 22 | 42.3 (37.1, 47.4) | −6.3 (−11.4, −1.2) |
| Lymphoma | 2554 | 49.8 (49.3, 50.3) | 40 | 47.2 (43.2, 51.2) | −2.6 (−6.5, 1.3) |
| Prostate | 13,528 | 51.0 (50.7, 51.3) | 78 | 43.0 (40.0, 46.1) | −8.0 (−11.1, −4.9) |
| VR-6Da, c | |||||
| Breast | 12,794 | 0.67 (0.67, 0.67) | 27 | 0.59 (0.55, 0.64) | −0.08 (−0.12, −0.04) |
| Colorectal | 5940 | 0.67 (0.67, 0.67) | 25 | 0.61 (0.56, 0.65) | −0.06 (−0.10, −0.02) |
| Gastrointestinal | 2338 | 0.64 (0.63, 0.64) | 99 | 0.60 (0.58, 0.63) | −0.04 (−0.06, −0.02) |
| Head and neck | 1799 | 0.66 (0.65, 0.66) | 19 | 0.59 (0.54, 0.64) | −0.07 (−0.11, −0.03) |
| Lung | 4681 | 0.63 (0.63, 0.64) | 22 | 0.57 (0.54, 0.61) | −0.06 (−0.10, −0.02) |
| Lymphoma | 2556 | 0.66 (0.66, 0.67) | 40 | 0.62 (0.58, 0.66) | −0.04 (−0.08, 0.00) |
| Prostate | 13,533 | 0.69 (0.68, 0.69) | 78 | 0.62 (0.60, 0.65) | −0.07 (−0.09, −0.05) |
All models were adjusted for age at first cancer diagnosis, sex, race/ethnicity, highest educational attainment, marital status, annual household income, cancer stage at diagnosis, whether the participant had multiple cancer diagnoses, whether the participant was diagnosed with any of 12 chronic conditions, months from diagnosis to survey, mode of survey administration, and whether the survey was completed by a proxy respondent.
Bolded Physical and Mental Component Summary scores represent a minimally important difference of 2.0 or greater between the HIV-negative and HIV-positive groups.
Bolded VR-6D scores represent a minimally important difference of 0.03 or greater between the HIV-negative and HIV-positive groups.
Adjusting for covariates, the mean difference in at least one HRQoL outcome exceeded the MID (2 points or greater for PCS and MCS; 0.03 or greater for VR-6D) for all cancer sites examined (Fig. 1). Differences in the PCS, MCS, and VR-6D scores for PWH compared to patients without HIV exceeded the MID for cancers of the breast (PCS: −2.7, 95% CI: −7.4, 2.0; MCS: −9.1, 95% CI: −14.2, −4.0; VR-6D: −0.08, 95% CI: −0.12, −0.04), colorectum (PCS: −2.3, 95% CI: −7.0, 2.4; MCS: −6.2, 95% CI: −10.7, −1.7; VR-6D: −0.06, 95% CI: −0.10, −0.02), gastrointestinal tract (MCS: −5.9, 95% CI: −8.7, −3.1; VR-6D: −0.04, 95% CI: −0.06, −0.02), head and neck (PCS: −4.1, 95% CI: −8.1, −0.1; MCS: −7.5, 95% CI: −13.2, −1.8; VR-6D: −0.07, 95% CI: −0.11, −0.03), lung (PCS: 2.5, 95% CI: −6.6, 1.6; MCS: −6.3, 95% CI: −11.4, −1.2; VR-6D: −0.06, 95% CI: −0.10, −0.02), lymphatic system (PCS: −2.3, 95% CI: −6.8, 2.2; MCS: −2.6, 95% CI: −6.5, 1.3; VR-6D: −0.04, 95% CI: −0.08, 0.00), and prostate (MCS: −8.0, 95% CI: −11.1, −4.9; VR-6D: −0.07, 95% CI: −0.09, −0.05) (Fig. 1).
Fig. 1.
Adjusted mean differences in a) Physical and Mental Component Summary scores and b) Veterans RAND 6-Dimension Health Survey (VR-6D) utility scores between HIV-positive and HIV-negative patients by cancer sites.
Similar findings between patients with and without HIV were observed for the eight VR-12 scale scores (Table 4). Physical functioning scores were similar between patients with and without HIV for colorectal, gastrointestinal, lung, lymphoma, and prostate cancers (Fig. 2). The largest differences in physical functioning scores for PWH compared to patients without HIV were for patients with head and neck (−3.5, 95% CI: −8.8, 1.8) and breast (−3.5, 95% CI: −9.6, 2.6) cancers, where the MID of 3 points was exceeded. Among PWH, physical functioning score was lowest for patients with lung cancer (31.9) (Table 4). Scores for role limitations due to physical health problems were similar between patients with and without HIV for breast, gastrointestinal, lung, lymphoma, and prostate cancers (Fig. 2). Differences in role physical scores between patients with and without HIV were found for patients with head and neck (−6.7, 95% CI: −11.2, −2.2) and colorectal (−3.3, 95% CI: −8.0, 1.4) cancers. Among PWH, patients with head and neck cancers reported the lowest role physical score (32.2) (Table 4). Bodily pain scores were similar between patients with and without HIV for gastrointestinal cancer (Fig. 2). PWH had lower bodily pain scores compared to patients without HIV for breast (−7.4, 95% CI: −12.5, −2.3), lung (−7.2, 95% CI: −11.1, −3.3), lymphoma (−4.3, 95% CI: −7.8, −0.8), colorectal (−4.2, 95% CI: −9.1, 0.7), prostate (−3.8, 95% CI: −6.4, −1.2), and head and neck (−3.2, 95% CI: −7.2, 0.8) cancers. Among PWH, patients with lung cancer had the lowest bodily pain score (33.0) (Table 4). General health scores were similar for patients with and without HIV for gastrointestinal, lung, and lymphoma cancers (Fig. 2). PWH reported lower general health scores for head and neck (−8.3, 95% CI: −14.2, −2.4), colorectal (−5.5, 95% CI: −10.8, −0.2), breast (−4.0, 95% CI: −10.1, 2.1), and prostate (−3.9, 95% CI: −7.0, −0.8) cancers. Among PWH, patients with head and neck cancer had the lowest general health score (32.2) (Table 4).
Table 4.
Adjusted mean Veterans RAND 12-Item Health Survey (VR-12) scale scores by HIV status and cancer sites.
| Cancer | HIV negative | HIV positive | Difference (95% CI) | ||
|---|---|---|---|---|---|
| N | Mean (95% CI) | N | Mean (95% CI) | ||
| Physical functioninga | |||||
| Breast | 12,695 | 37.3 (37.0, 37.6) | 27 | 33.8 (27.7, 39.9) | −3.5 (−9.6, 2.6) |
| Colorectal | 5884 | 37.2 (36.9, 37.5) | 25 | 34.4 (29.1, 39.6) | −2.8 (−8.1, 2.5) |
| Gastrointestinal | 2319 | 34.7 (34.1, 35.2) | 99 | 34.7 (31.8, 37.5) | 0.0 (−3.0, 3.0) |
| Head and neck | 1785 | 36.8 (36.1, 37.4) | 19 | 33.3 (27.9, 38.7) | −3.5 (−8.8, 1.8) |
| Lung | 4643 | 32.4 (32.0, 32.8) | 21 | 31.9 (26.6, 37.2) | −0.5 (−5.8, 4.8) |
| Lymphoma | 2549 | 36.3 (35.7, 36.8) | 39 | 34.1 (28.7, 39.5) | −2.2 (−7.7, 3.3) |
| Prostate | 13,445 | 39.2 (38.9, 39.5) | 78 | 37.3 (34.2, 40.4) | −1.9 (−5.1, 1.3) |
| Role - Physicala | |||||
| Breast | 12,680 | 39.7 (39.4, 40.0) | 25 | 37.1 (31.7, 42.5) | −2.6 (−8.1, 2.9) |
| Colorectal | 5888 | 39.4 (39.1, 39.7) | 25 | 36.1 (31.5, 40.7) | −3.3 (−8.0, 1.4) |
| Gastrointestinal | 2314 | 36.7 (36.2, 37.2) | 98 | 36.8 (34.4, 39.3) | 0.1 (−2.5, 2.7) |
| Head and neck | 1779 | 38.9 (38.3, 39.4) | 19 | 32.2 (27.6, 36.7) | −6.7 (−11.2, −2.2) |
| Lung | 4640 | 36.1 (35.7, 36.4) | 22 | 33.2 (28.7, 37.6) | −2.9 (−7.4, 1.6) |
| Lymphoma | 2543 | 38.4 (37.9, 38.9) | 39 | 35.8 (31.5, 40.2) | −2.6 (−6.9, 1.7) |
| Prostate | 13,420 | 40.9 (40.6, 41.2) | 77 | 38.6 (35.6, 41.7) | −2.3 (−5.2, 0.6) |
| Bodily paina | |||||
| Breast | 12,611 | 41.8 (41.6, 42.1) | 27 | 34.4 (29.4, 39.4) | −7.4 (−12.5, −2.3) |
| Colorectal | 5831 | 42.1 (41.8, 42.4) | 24 | 37.9 (32.9, 42.8) | −4.2 (−9.1, 0.7) |
| Gastrointestinal | 2290 | 40.3 (39.8, 40.8) | 99 | 38.0 (35.8, 40.1) | −2.3 (−4.5, −0.1) |
| Head and neck | 1768 | 41.1 (40.6, 41.6) | 18 | 37.9 (34.0, 41.7) | −3.2 (−7.2, 0.8) |
| Lung | 4596 | 40.2 (39.8, 40.5) | 21 | 33.0 (29.2, 36.8) | −7.2 (−11.1, −3.3) |
| Lymphoma | 2517 | 41.5 (41.0, 41.9) | 39 | 37.2 (33.7, 40.7) | −4.3 (−7.8, −0.8) |
| Prostate | 13,322 | 42.7 (42.4, 43.0) | 77 | 38.9 (36.3, 41.4) | −3.8 (−6.4, −1.2) |
| General healtha | |||||
| Breast | 12,557 | 41.9 (41.6, 42.2) | 24 | 37.9 (31.8, 44.0) | −4.0 (−10.1, 2.1) |
| Colorectal | 5808 | 41.3 (41.0, 41.6) | 24 | 35.8 (30.5, 41.2) | −5.5 (−10.8, −0.2) |
| Gastrointestinal | 2297 | 37.3 (36.8, 37.8) | 97 | 37.6 (34.7, 40.4) | 0.3 (−2.7, 3.3) |
| Head and neck | 1762 | 40.5 (39.9, 41.0) | 19 | 32.2 (26.2, 38.1) | −8.3 (−14.2, −2.4) |
| Lung | 4587 | 37.2 (36.8, 37.6) | 21 | 34.7 (29.2, 40.3) | −2.5 (−8.0, 3.0) |
| Lymphoma | 2513 | 41.0 (40.5, 41.5) | 40 | 39.0 (34.2, 43.8) | −2.0 (−6.7, 2.7) |
| Prostate | 13,359 | 42.4 (42.1, 42.7) | 77 | 38.5 (35.3, 41.6) | −3.9 (−7.0, −0.8) |
| Vitalitya | |||||
| Breast | 12,548 | 45.4 (45.2, 45.7) | 26 | 40.0 (35.4, 44.7) | −5.4 (−10.1, −0.7) |
| Colorectal | 5818 | 45.4 (45.1, 45.7) | 25 | 44.4 (40.1, 48.7) | −1.0 (−5.3, 3.3) |
| Gastrointestinal | 2295 | 43.0 (42.5, 43.5) | 99 | 42.1 (39.8, 44.5) | −0.9 (−3.3, 1.5) |
| Head and neck | 1753 | 44.8 (44.2, 45.3) | 19 | 41.5 (36.6, 46.4) | −3.3 (−8.2, 1.6) |
| Lung | 4592 | 42.3 (41.9, 42.7) | 22 | 36.7 (32.2, 41.1) | −5.6 (−10.1, −1.1) |
| Lymphoma | 2505 | 44.9 (44.4, 45.4) | 39 | 45.2 (41.1, 49.3) | 0.3 (−3.8, 4.4) |
| Prostate | 13,293 | 46.9 (46.6, 47.2) | 77 | 43.6 (40.4, 46.7) | −3.3 (−6.4, −0.2) |
| Social functioninga | |||||
| Breast | 12,565 | 44.1 (43.8, 44.5) | 26 | 34.5 (28.2, 40.7) | −9.6 (−15.9, −3.3) |
| Colorectal | 5847 | 43.7 (43.4, 44.1) | 24 | 37.8 (33.0, 42.6) | −5.9 (−10.6, −1.2) |
| Gastrointestinal | 2296 | 40.6 (40.0, 41.1) | 99 | 34.5 (31.6, 37.4) | −6.1 (−9.1, −3.1) |
| Head and neck | 1762 | 42.7 (42.1, 43.4) | 19 | 33.8 (28.8, 38.9) | −8.9 (−14.0, −3.8) |
| Lung | 4603 | 41.0 (40.6, 41.4) | 21 | 32.3 (25.8, 38.7) | −8.7 (−15.2, −2.2) |
| Lymphoma | 2519 | 43.2 (42.7, 43.8) | 40 | 40.1 (35.5, 44.8) | −3.1 (−7.8, 1.6) |
| Prostate | 13,343 | 45.3 (45.0, 45.6) | 78 | 37.2 (33.8, 40.6) | −8.1 (−11.5, −4.7) |
| Role - Emotionala | |||||
| Breast | 12,676 | 46.3 (46.0, 46.6) | 26 | 37.9 (32.1, 43.7) | −8.4 (−14.3, −2.5) |
| Colorectal | 5875 | 46.1 (45.8, 46.4) | 25 | 38.5 (33.8, 43.3) | −7.6 (−12.3, −2.9) |
| Gastrointestinal | 2312 | 43.9 (43.4, 44.4) | 97 | 38.9 (36.1, 41.7) | −5.0 (−7.8, −2.2) |
| Head and neck | 1776 | 45.4 (44.8, 45.9) | 19 | 37.8 (31.9, 43.7) | −7.6 (−13.5, −1.7) |
| Lung | 4621 | 44.3 (43.9, 44.7) | 22 | 39.8 (33.6, 46.0) | −4.5 (−10.8, 1.8) |
| Lymphoma | 2533 | 45.5 (45.0, 46.0) | 40 | 43.8 (39.0, 48.7) | −1.7 (−6.6, 3.2) |
| Prostate | 13,417 | 47.2 (46.9, 47.4) | 77 | 39.9 (36.7, 43.1) | −7.3 (−10.4, −4.2) |
| Mental healtha | |||||
| Breast | 12,659 | 48.7 (48.5, 49.0) | 26 | 41.1 (35.6, 46.7) | −7.6 (−13.1, −2.1) |
| Colorectal | 5880 | 48.7 (48.3, 49.0) | 24 | 42.1 (37.2, 47.1) | −6.6 (−11.5, −1.7) |
| Gastrointestinal | 2310 | 46.6 (46.1, 47.2) | 98 | 42.5 (39.7, 45.3) | −4.1 (−6.9, −1.3) |
| Head and neck | 1770 | 47.3 (46.7, 47.9) | 19 | 40.5 (34.6, 46.4) | −6.8 (−12.7, −0.9) |
| Lung | 4625 | 46.8 (46.5, 47.2) | 22 | 42.8 (37.9, 47.6) | −4.0 (−8.9, 0.9) |
| Lymphoma | 2526 | 48.6 (48.1, 49.1) | 40 | 43.5 (39.4, 47.7) | −5.1 (−9.3, −0.9) |
| Prostate | 13,413 | 49.7 (49.4, 50.0) | 78 | 43.3 (40.3, 46.3) | −6.4 (−9.3, −3.5) |
All models were adjusted for age at first cancer diagnosis, sex, race/ethnicity, highest educational attainment, marital status, annual household income, cancer stage at diagnosis, whether the participant had multiple cancer diagnoses, whether the participant was diagnosed with any of 12 chronic conditions, months from diagnosis to survey, mode of survey administration, and whether the survey was completed by a proxy respondent. Bolded VR-12 scale scores represent a minimally important difference of 3.0 or greater between the HIV-negative and HIV-positive groups.
Fig. 2.
Adjusted mean differences in Veterans RAND 12-Item Health Survey (VR-12) scale scores between HIV-positive and HIV-negative patients by cancer sites.
PWH with colorectal, gastrointestinal, and lymphoma cancers reported similar vitality scores compared to patients without HIV (Fig. 2). PWH had lower vitality score compared to patients without HIV for lung (−5.6, 95% CI: −10.1, −1.1), breast (−5.4, 95% CI: −10.1, −0.7), head and neck (−3.3, 95% CI: −8.2, 1.6), and prostate (−3.3, 95% CI: −6.4, −0.2) cancers. PWH with lung cancer reported the lowest vitality score (36.7) (Table 4). Compared to patients without HIV, PWH had lower social functioning scores across all cancers, including breast (−9.6, 95% CI: −15.9, −3.3), head and neck (−8.9, 95% CI: −14.0, −3.8), lung (−8.7, 95% CI: −15.2, −2.2), prostate (−8.1, 95% CI: −11.5, −4.7), gastrointestinal (−6.1, 95% CI: −9.1, −3.1), colorectal (−5.9, 95% CI: −10.6, −1.2), and lymphoma (−3.1, 95% CI: −7.8, 1.6) cancers (Fig. 2). PWH with lung cancer reported the lowest social functioning scores (32.3) (Table 4). Compared to patients without HIV, PWH had lower scores for role limitations due to personal or emotional problems across all cancers except lymphoma, including breast (−8.4, 95% CI: −14.3, −2.5), head and neck (−7.6, 95% CI: −13.5, −1.7), colorectal (−7.6, 95% CI: −12.3, −2.9), prostate (−7.3, 95% CI: −10.4, −4.2), gastrointestinal (−5.0, 95% CI: −7.8, −2.2), and lung (−4.5, 95% CI: −10.8, 1.8) cancers (Fig. 2). PWH with head and neck (37.8) and breast (37.9) cancers reported the lowest role emotional scores (Table 4). PWH across all cancers reported lower mental health scores compared to patients without HIV, including breast (−7.6, 95% CI: −13.1, −2.1), head and neck (−6.8, 95% CI: −12.7, −0.9), colorectal (−6.6, 95% CI: −11.5, −1.7), prostate (−6.4, 95% CI: −9.3, −3.5), lymphoma (−5.1, 95% CI: −9.3, −0.9), gastrointestinal (−4.1, 95% CI: −6.9, −1.3), and lung (−4.0, 95% CI: −8.9, 0.9) cancers (Fig. 2). Among PWH, patients with head and neck cancer reported the lowest mental health score (40.5) (Table 4). PWH with head and neck cancer reported lower HRQoL on all eight scales compared to patients without HIV.
4. Discussion
This study aimed to assess differences in HRQoL among a national sample of adults with non-AIDS-defining cancers with and without HIV. To our knowledge, limited research exists examining patient-reported HRQoL among PWH with cancer. Our findings demonstrated that PWH with cancer experience substantially lower HRQoL across all the cancers examined. In particular, PWH with breast, colorectal, and head and neck cancers exhibited the greatest difference in HRQoL compared to patients without HIV. Future work can explore strategies for monitoring and managing symptoms among PWH with cancer, particularly among older patients. The inclusion of Medicare-eligible PWH under age 65 in our study played a critical role in assessing the difference in symptom burden for PWH given the overall aging of people with HIV and the growing population of older adults with HIV with cancer [5]. Leveraging the SEER-Medicare data has been critical to assessing the differences in cancer-related outcomes among people with and without HIV in prior seminal work [40–42]. Compared to people without HIV, PWH experience geriatric conditions at a younger age, including living with a greater number of comorbidities [43], functional impairments (particularly due to frailty) [44], cognitive decline [43], physical function decline [45], and impairments in activities of daily living [43]. The greater prevalence of aging-related conditions in younger PWH due to advanced HIV disease can severely impact outcomes across the cancer care continuum, including cancer treatment receipt, care quality, and mortality [2,6–10].
In cancer care, patients experience multifaceted symptoms due to the side effects associated with systemic therapy and treatments generally [15]. High symptom burden associated with cancer treatment is known to lead to lower patient satisfaction, adherence to cancer treatment, and survival probabilities [14,46,47]. Living with comorbidities while undergoing cancer treatment exacerbates existing symptoms [12,48]. This is particularly critical for the case of PWH with cancer given the symptomology associated with living with HIV and adherence to ART medications. Long-term use of ART to control HIV is associated with digestive discomfort, skin rashes, numbness, memory loss, dizziness, as well as a high burden of depressive symptoms [49]. Additionally, PWH can also face clinical factors that contribute to lower HRQoL, including poor HIV control and high HIV viral load. Lower CD4 cell count is associated with worse physical health and greater disease progression [50,51]. Given the dual burden of living with both HIV and cancer, future work can examine interventions and care coordination models to monitor patient-reported outcomes and manage symptom burden among PWH with cancer.
Our results are similar to one prior study that evaluated patient-reported outcomes among PWH with cancer [52]. Evaluated using the Edmonton Symptom Assessment Scale (ESAS), symptom burden, including pain, anxiety, depression, and nausea, was significantly higher among PWH compared to those without HIV [52]. Although to our knowledge, symptom burden and HRQoL have not been previously evaluated among PWH with cancers besides head and neck, there is extensive literature demonstrating that patients across many cancer types experience lower HRQoL compared with the general population [12,16,17]. In our study, among the cancers examined, PWH with head and neck cancers reported some of the lowest PCS, MCS, and VR-6D scores. We also found that PWH with head and neck cancers scored the lowest on HRQoL physical scales (physical functioning, role physical, and general health) and mental scales (role emotional and mental health). This is consistent with the symptom profile and treatment-associated side effects of head and neck cancers, including the loss of functions for swallowing, eating, and speaking [53].
Addressing high symptom burden and improving HRQoL is a major priority for cancer care among older patients. Older cancer patients face unique care needs but despite this, a review found that 15–93% of older patients undergoing cancer treatment had unmet needs [54]. The most common needs included psychological and emotional (e.g., anxiety, fear), need for information, and physical (e.g., fatigue, pain) [54,55]. These needs are most prevalent immediately after diagnosis and during treatment [54]. Geriatric assessments to detect and monitor unaddressed needs should be used to inform care plans that incorporate patient preferences and shared decision-making [56]. In 2014, the National Comprehensive Cancer Network recommended that palliative care should be integrated into routine cancer care at the start of diagnosis, particularly for high-risk patients with competing comorbidities [57]. Palliative care provides an opportunity for patients to discuss symptom burden with a dedicated care team focused on mitigating the adverse impacts of symptoms during the cancer treatment experience. However, despite evidence showing that receipt of palliative care early in cancer care leads to improved HRQoL and survival [58–60], PWH are less likely to access palliative or supportive care services, particularly PWH of low socioeconomic backgrounds [8–10]. Access to supportive care services is influenced by multilevel individual, healthcare system, and social level factors [61]. In addition, PWH face unique barriers to cancer care such as low social support, stigma, and discrimination due to their HIV status [50]. These social factors negatively impact the receipt and quality of cancer care that PWH receive in the US [50]. Integrated palliative care interventions can reduce social barriers to care and HIV stigma among PWH [62].
5. Limitations
This study should be considered in light of the following limitations. First, we were only able to determine a participant’s HIV status based on their ART medication use due to the lack of information on HIV laboratory tests and diagnosis codes in the SEER-MHOS linkage data. Ideally, HIV-positive status would be determined based on validated approaches that used a combination of data on HIV medication, tests, and diagnosis codes [22]. Furthermore, the small number of participants with HIV may limit the precision of the estimates and generalizability of the results to the population of PWH. However, we observed substantial disparities in HRQoL between participants with and without HIV even though the potential misclassification of HIV status would bias our results toward the null. Second, the SEER data only included Medicare Advantage beneficiaries from 12 participating states and 4 metropolitan areas [63]. Therefore, findings from this study may not be representative of Medicare fee-for-service beneficiaries or of patients from non-participating geographic areas. Third, only services covered by Medicare were available in the SEER-MHOS data. HIV-related clinical information, such as viral load and CD4 cell count, were also not available. Receipt of cancer care not covered by Medicare and HIV-related clinical information that may be important in explaining a participant’s HRQoL were not accounted for in this study. Fourth, the time from cancer diagnosis to survey completion differed by HIV status and cancer type, even though it was adjusted for in the analysis. Ideally, a future dataset would have a measurement of HRQoL taken close to active cancer treatment to better attribute the effect of HRQoL on patient cancer outcomes, such as treatment tolerance and adherence. Fifth, there is the potential for non-response bias to the MHOS based on certain sociodemographic characteristics including age, race/ethnicity, sex, and whether patient was residing in a long-term institutionalized facility [64,65]. Furthermore, the non-response bias among people with and without HIV is unknown. However, prior studies have found that the non-response bias is modest and did not bias estimations of health status or cause the resulting sample to be non-representative of the eligible population [64–66].
6. Conclusion
Our findings demonstrated that PWH across all cancers examined reported substantially lower mental and/or physical HRQoL compared to patients without HIV. In particular, PWH with breast, colorectal, and head and neck cancers exhibited the greatest difference in HRQoL compared to patients without HIV. Among PWH, patients with head and neck cancers reported some of the lowest PCS, MCS, and VR-6D scores. Future work can explore strategies for symptom monitoring and management to identify and address high symptom burden among PWH.
Supplementary Material
Acknowledgements
This study used data from the SEER-MHOS linked data resource. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the National Cancer Institute; the Centers for Medicare & Medicaid Services; Information Management Services (IMS), Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-MHOS database. This work has been supported in part by the Biostatistics and Bioinformatics Shared Resource at the H. Lee Moffitt Cancer Center & Research Institute, an NCI-Designated Comprehensive Cancer Center.
Funding
This study was funded by the National Cancer Institute (grant R03 CA278603-01A1).
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Jessica Y. Islam reports financial support was provided by National Cancer Institute. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jgo.2026.102855.
Footnotes
CRediT authorship contribution statement
Yu Chen Lin: Methodology, Formal analysis, Writing – original draft. Di Kang: Methodology, Formal analysis, Writing – review & editing. Biwei Cao: Methodology, Formal analysis, Writing – review & editing. Emma Hume: Project administration, Writing – review & editing. Amir Alishahi Tabriz: Writing – review & editing. Gita Suneja: Writing – review & editing. Anna E. Coghill: Writing – review & editing. Heather Jim: Writing – review & editing. Kea Turner: Writing – review & editing. Jessica Y. Islam: Conceptualization, Methodology, Formal analysis, Supervision, Project administration, Writing – review & editing.
Ethics approval
This study was reviewed and deemed exempt from formal review by the Moffitt Cancer Center Institutional Review Board of Record, Advarra. Participants of the MHOS survey provided informed consent. The SEER-MHOS linkage data are considered a limited dataset with no personal identifying information and is therefore exempt from additional informed consent requirements.
Data availability
The data analyzed in this study are available from SEER-MHOS. Restrictions apply to the availability of these data, and investigators interested in using the SEER-MHOS data must submit a request for access.
References
- [1].Shiels MS, Pfeiffer RM, Gail MH, Hall HI, Li J, Chaturvedi AK, et al. Cancer burden in the HIV-infected population in the United States. J Natl Cancer Inst 2011;103(9): 753–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Coghill AE, Pfeiffer RM, Shiels MS, Engels EA. Excess mortality among HIV-infected individuals with cancer in the United States. Cancer Epidemiol Biomarkers Prev 2017;26(7):1027–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].HIV-Causal Collaboration. The effect of combined antiretroviral therapy on the overall mortality of HIV-infected individuals. AIDS 2010;24(1):123–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Robbins HA, Shiels MS, Pfeiffer RM, Engels EA. Epidemiologic contributions to recent cancer trends among HIV-infected people in the United States. AIDS 2014;28 (6):881–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Shiels MS, Islam JY, Rosenberg PS, Hall HI, Jacobson E, Engels EA. Projected cancer incidence rates and burden of incident cancer cases in HIV-infected adults in the United States through 2030. Ann Intern Med 2018;168(12):866–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Suneja G, Shiels MS, Angulo R, Copeland GE, Gonsalves L, Hakenewerth AM, et al. Cancer treatment disparities in HIV-infected individuals in the United States. J Clin Oncol 2014;32(22):2344–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].McGee-Avila JK, Suneja G, Engels EA, Rositch AF, Horner M-J, Luo Q, et al. Cancer treatment disparities in people with HIV in the United States, 2001–2019. J Clin Oncol 2024;42(15):1810–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Islam JY, Nogueira L, Suneja G, Coghill A, Akinyemiju T. Palliative care use among people living with HIV and cancer: an analysis of the National Cancer Database (2004–2018). JCO Oncol Pract 2022;18(10):e1683–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Islam JY, Guo Y, Turner K, Tabriz AA, Lin YC, Vidot DC, et al. Inequities in palliative care delivery to patients with HIV and stage IV cancers in the US (2004–2020). JNCI Cancer Spectrum 2024;9(2):pkae118. [Google Scholar]
- [10].Ahmad JN, Islam JY, Aboulafia DM. Supportive care for people living with HIV and cancer in the United States. Int. J. Cancer Care Deliv 2024;4(1). [Google Scholar]
- [11].Andresen EM, Meyers AR. Health-related quality of life outcomes measures. Arch Phys Med Rehabil 2000;81:S30–45. [DOI] [PubMed] [Google Scholar]
- [12].Smith AW, Reeve BB, Bellizzi KM, Harlan LC, Klabunde CN, Amsellem M, et al. Cancer, comorbidities, and health-related quality of life of older adults. Health Care Financ Rev 2008;29(4):41. [PMC free article] [PubMed] [Google Scholar]
- [13].Emuren L, Welles S, Polansky M, Evans AA, Macalino G, Agan BK. Lower health-related quality of life predicts all-cause hospitalization among HIV-infected individuals. Health Qual Life Outcomes 2018;16:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Brown DS, Thompson WW, Zack MM, Arnold SE, Barile JP. Associations between health-related quality of life and mortality in older adults. Prev Sci 2015;16:21–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Henry DH, Viswanathan HN, Elkin EP, Traina S, Wade S, Cella D. Symptoms and treatment burden associated with cancer treatment: results from a cross-sectional national survey in the US. Support Care Cancer 2008;16:791–801. [DOI] [PubMed] [Google Scholar]
- [16].Kent EE, Ambs A, Mitchell SA, Clauser SB, Smith AW, Hays RD. Health-related quality of life in older adult survivors of selected cancers: data from the SEER-MHOS linkage. Cancer 2015;121(5):758–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Baumann R, Pütz C, Röhrig B, Höffken K, Wedding U. Health-related quality of life in elderly cancer patients, elderly non-cancer patients and an elderly general population. Eur J Cancer Care 2009;18(5):457–65. [Google Scholar]
- [18].National Cancer Institute. SEER-medicare health outcomes survey (SEER-MHOS) linked data resource. https://healthcaredelivery.cancer.gov/seer-mhos/; 2025.
- [19].KFF. Medicare and people with HIV. https://www.kff.org/hivaids/issue-brief/medicare-and-people-with-hiv/; March 27, 2023.
- [20].Health Services Advisory Group. Medicare HOS survey status information. https://www.hosonline.org/globalassets/hos-online/survey-results/mhos_survey_status_information_c22.pdf; August, 2022.
- [21].Surveillance, Epidemiology, and End Results Program. Site recode ICD-O-3/WHO 2008 definition. https://seer.cancer.gov/siterecode/icdo3_dwhoheme/index.html; 2025.
- [22].Ridgway JP, Mason JA, Friedman EE, Devlin S, Zhou J, Meltzer D, et al. Comparison of algorithms for identifying people with HIV from electronic medical records in a large, multi-site database. JAMIA Open 2022;5(2):ooac033. [Google Scholar]
- [23].Centers for Disease Control and Prevention. U.S. Department of Health and Human Services. Updated guidelines for antiretroviral postexposure prophylaxis after sexual, injection drug use, or other nonoccupational exposure to HIV—United States, 2016. https://www.cdc.gov/hiv/pdf/programresources/cdc-hiv-npep-guidelines.pdf; 2016.
- [24].Framework IC, The MOS. 36-item short-form health survey (SF-36). Med Care 1992;30(6):473–83. [PubMed] [Google Scholar]
- [25].Ware J, Kosinski M, Keller S. SF-36 physical and mental health summary scales: A user’s manual. Health Assessment Lab; 1994. [Google Scholar]
- [26].Hays RD, Farivar SS, Liu H. Approaches and recommendations for estimating minimally important differences for health-related quality of life measures. COPD: J Chron Obstruct Pulmon Dis 2005;2(1):63–7. [Google Scholar]
- [27].Boston University School of Public Health. VR-36, VR-12 and VR-6D. https://www.bu.edu/sph/research/centers-and-groups/vr-36-vr-12-and-vr-6d/; 2025.
- [28].Selim AJ, Rogers W, Qian SX, Brazier J, Kazis LE. A preference-based measure of health: the VR-6D derived from the veterans RAND 12-Item Health Survey. Qual Life Res 2011;20:1337–47. [DOI] [PubMed] [Google Scholar]
- [29].Gandhi PK, Ried LD, Bibbey A, Huang I-C. SF-6D utility index as measure of minimally important difference in health status change. J Am Pharm Assoc 2012; 52(1):34–42. [Google Scholar]
- [30].Jia H, Lubetkin EI, DeMichele K, Stark DS, Zack MM, Thompson WW. Comparing the performance of 2 health utility measures in the Medicare health outcome survey (HOS). Med Decis Making 2018;38(8):983–93. [DOI] [PubMed] [Google Scholar]
- [31].Graubard BI, Korn EL. Predictive margins with survey data. Biometrics 1999;55(2): 652–9. [DOI] [PubMed] [Google Scholar]
- [32].Lumley T, Scott A. Fitting regression models to survey data. Stat Sci 2017:265–78. [Google Scholar]
- [33].Hays RD, Spritzer KL. REcycled SAS® PrEdiCTions (RESPECT). https://labs.dgsom.ucla.edu/hays/files/view/docs/programs-utilities/HaysRecycle.pdf; 2013.
- [34].Ahmed ES, Butt MN. The misunderstood P-value: Why statistical significance is not enough in clinical practice. Elsevier; 2025. [Google Scholar]
- [35].Sedaghat AR. Understanding the minimal clinically important difference (MCID) of patient-reported outcome measures. Otolaryngology 2019;161(4):551–60. [Google Scholar]
- [36].Rosenbaum PR, Rubin DB. Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. Am Stat 1985;39 (1):33–8. [Google Scholar]
- [37].Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika 1983;70(1):41–55. [Google Scholar]
- [38].Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivar Behav Res 2011;46(3):399–424. [Google Scholar]
- [39].National Cancer Institute. Propensity score matching guidance. https://healthcaredelivery.cancer.gov/seer-mhos/program/propensity-score-matching.html; February 24, 2025.
- [40].Coghill AE, Suneja G, Rositch AF, Shiels MS, Engels EA. HIV infection, cancer treatment regimens, and cancer outcomes among elderly adults in the United States. JAMA Oncol 2019;5(9):e191742. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Islam JY, Zhou X, Baggett C, Jamjian C, Coghill A, Suneja G, et al. Potential drug-drug interactions between cancer therapies and active antiretroviral treatments among Medicare-enrolled patients with cancer and HIV. J Geriatr Oncol 2025;16 (2):102158. [DOI] [PubMed] [Google Scholar]
- [42].Rositch AF, Jiang S, Coghill AE, Suneja G, Engels EA. Disparities and determinants of cancer treatment in elderly Americans living with human immunodeficiency virus/AIDS. Clin Infect Dis 2018;67(12):1904–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [43].Vance DE, Fazeli PL, Gakumo CA. The impact of neuropsychological performance on everyday functioning between older and younger adults with and without HIV. J Assoc Nurs AIDS Care 2013;24(2):112–25. [Google Scholar]
- [44].Desquilbet L, Jacobson LP, Fried LP, Phair JP, Jamieson BD, Holloway M, et al. HIV-1 infection is associated with an earlier occurrence of a phenotype related to frailty. J Gerontol A Biol Sci Med Sci 2007;62(11):1279–86. [DOI] [PubMed] [Google Scholar]
- [45].Oursler KK, Goulet JL, Crystal S, Justice AC, Crothers K, Butt AA, et al. Association of age and comorbidity with physical function in HIV-infected and uninfected patients: results from the Veterans Aging Cohort Study. AIDS Patient Care STDS 2011;25(1):13–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [46].Miller JL, Evers J. Barriers to adherence to cancer treatments among head and neck cancer patients. J Adv Pract Oncol 2022;13(5):515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [47].Qian CL, Kaslow-Zieve ER, Azoba CC, Horick N, Wang I, Van Seventer E, et al. Associations of patient-reported care satisfaction with symptom burden and healthcare use in hospitalized patients with cancer. Support Care Cancer 2022;30 (5):4527–36. [DOI] [PubMed] [Google Scholar]
- [48].Han X, Robinson LA, Jensen RE, Smith TG, Yabroff KR. Factors associated with health-related quality of life among cancer survivors in the United States. JNCI Cancer Spectrum 2021;5(1):pkaa123. [Google Scholar]
- [49].Chen W-T, Shiu C-S, Yang JP, Simoni JM, Lee TS-H, Zhao H. Antiretroviral therapy (ART) side effect impacted on quality of life, and depressive symptomatology: a mixed-method study. J AIDS Clin Res 2013;4:218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [50].Degroote S, Vogelaers D, Vandijck DM. What determines health-related quality of life among people living with HIV: an updated review of the literature. Archiv Public Health 2014;72:1–10. [Google Scholar]
- [51].Murri R, Fantoni M, Del Borgo C, Visona R, Barracco A, Zambelli A, et al. Determinants of health-related quality of life in HIV-infected patients. AIDS Care 2003;15(4):581–90. [DOI] [PubMed] [Google Scholar]
- [52].Coghill AE, Brownstein NC, Sinha S, Thompson ZJ, Dickey BL, Hoogland AI, et al. Patient-reported outcomes in cancer patients with HIV. Cancers 2022;14(23):5889. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [53].Hammermüller C, Hinz A, Dietz A, Wichmann G, Pirlich M, Berger T, et al. Depression, anxiety, fatigue, and quality of life in a large sample of patients suffering from head and neck cancer in comparison with the general population. BMC Cancer 2021;21:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [54].Puts M, Papoutsis A, Springall E, Tourangeau A. A systematic review of unmet needs of newly diagnosed older cancer patients undergoing active cancer treatment. Support Care Cancer 2012;20:1377–94. [DOI] [PubMed] [Google Scholar]
- [55].Thewes B, Butow P, Girgis A, Pendlebury S. The psychosocial needs of breast cancer survivors; a qualitative study of the shared and unique needs of younger versus older survivors. Psycho-Oncology 2004;13(3):177–89. [DOI] [PubMed] [Google Scholar]
- [56].Shahrokni A, Kim SJ, Bosl GJ, Korc-Grodzicki B. How we care for an older patient with cancer. J Oncol Pract 2017;13(2):95–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [57].Levy MH, Smith T, Alvarez-Perez A, Back A, Baker JN, Block S, et al. Palliative care, version 1.2014. J Natl Compr Canc Netw 2014;12(10):1379–88. [DOI] [PubMed] [Google Scholar]
- [58].Bakitas MA, Tosteson TD, Li Z, Lyons KD, Hull JG, Li Z, et al. Early versus delayed initiation of concurrent palliative oncology care: patient outcomes in the ENABLE III randomized controlled trial. J Clin Oncol 2015;33(13):1438–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [59].Bouleuc C, Burnod A, Angellier E, Massiani M-A, Robin M-L, Copel L, et al. Early palliative care in oncology. Bull Cancer 2019;106(9):796–804. [DOI] [PubMed] [Google Scholar]
- [60].Salins N, Ramanjulu R, Patra L, Deodhar J, Muckaden MA. Integration of early specialist palliative care in cancer care and patient related outcomes: a critical review of evidence. Indian J Palliat Care 2016;22(3):252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [61].Zubritsky C, Abbott KM, Hirschman KB, Bowles KH, Foust JB, Naylor MD. Health-related quality of life: expanding a conceptual framework to include older adults who receive long-term services and supports. Gerontologist 2013;53(2):205–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [62].Lowther K, Harding R, Simms V, Gikaara N, Ahmed A, Ali Z, et al. Effect of participation in a randomised controlled trial of an integrated palliative care intervention on HIV-associated stigma. AIDS Care 2018;30(9):1180–8. [DOI] [PubMed] [Google Scholar]
- [63].Office of Disease Prevention and Health Promotion. Surveillance, epidemiology, and end results program (SEER). https://odphp.health.gov/healthypeople/objectives-and-data/data-sources-and-methods/data-sources/surveillance-epidemiology-and-end-results-program-seer; 2025.
- [64].McCall N, Bonito A, Trofimovich L. Estimation and analysis of non-response bias in medicare surveys. RTI International; 2004. [RTI Project Number 08529.002]. [Google Scholar]
- [65].McCall N, Khatutsky G, Smith K, Pope GC. Estimation of non-response bias in the Medicare FFS HOS. Health Care Financ Rev 2004;25(4):27. [PMC free article] [PubMed] [Google Scholar]
- [66].Ambs A, Warren JL, Bellizzi KM, Topor M, Haffer SCC, Clauser SB. Overview of the SEER—medicare health outcomes survey linked dataset. Health Care Financ Rev 2008;29(4):5. [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data analyzed in this study are available from SEER-MHOS. Restrictions apply to the availability of these data, and investigators interested in using the SEER-MHOS data must submit a request for access.