Abstract
Background
Advanced chronic obstructive pulmonary disease (COPD) is associated with multimorbidity and substantial healthcare use. However, sex differences in documented multimorbidity and healthcare use near the end of life remain insufficiently understood.
Methods
This retrospective population-based cohort study used routinely collected administrative and clinical data from Clalit Health Services, Israel. The study included deceased patients with advanced COPD during 2016–2020 and 2022–2023. The year 2021 was excluded a priori as the period designated in the study protocol to minimize distortion in healthcare utilization, and mortality patterns; 2020 remained part of the prespecified observation window. The cohort comprised 2261 patients aged ≥55 years who met an operational definition requiring documented COPD, FEV1/FVC <0.70, FEV1 <50% predicted, and at least two opioid prescriptions or dispensations during the final year of life. The opioid criterion was not considered a stand-alone diagnostic criterion or evidence of refractory dyspnea. Women and men were compared regarding demographic characteristics, multimorbidity, selected healthcare-use indicators, coded palliative care encounters, and time from first documented COPD diagnosis to death.
Results
The cohort included 547 women and 1714 men. Women were older and more frequently had dementia and rheumatic/connective tissue disease. Men more frequently had myocardial infarction, renal disease, peripheral vascular disease, and diabetes. Men had more documented procedures, although the effect size was small. The mean Charlson Comorbidity Index differed by 0.44 points, of uncertain clinical importance. Coded palliative care encounters were rare and available only for small subsamples. Median time from first documented COPD diagnosis to death was 33 months among women and 29 months among men, without a statistically significant difference (log-rank p=0.328).
Conclusion
Sex differences were observed primarily in documented multimorbidity. Healthcare-use and palliative-care findings were partial and descriptive and should not be interpreted as complete end-of-life trajectories or causal effects of sex.
Keywords: chronic obstructive pulmonary disease, sex differences, multimorbidity, palliative care, end of life, retrospective cohort
Plain Language Summary
COPD is a long-term lung disease that may occur together with several other health conditions. These conditions can influence symptoms, treatment, and the use of health services. We examined whether documented illness patterns differed between women and men in a selected group of deceased patients with advanced COPD.
We used routinely collected data from Clalit Health Services in Israel. The study included 2261 deceased patients aged 55 years or older during 2016-2020 and 2022-2023. All patients had a documented COPD diagnosis, severe airflow limitation, and at least two opioid prescriptions or dispensations in the final year of life. Opioids may have been prescribed for breathlessness, pain, or another condition; therefore, this criterion does not prove that patients had refractory breathlessness or received palliative care.
Women were older and more often had dementia and rheumatic or connective tissue disease. Men more often had myocardial infarction, renal disease, peripheral vascular disease, and diabetes. Men also had more coded procedures, but the difference was small. Formally coded palliative-care encounters were uncommon and available for only small groups. Median time from first documented COPD diagnosis to death was 33 months for women and 29 months for men, with no statistically significant difference.
The clearest findings concerned different patterns of documented multimorbidity. The healthcare-use and palliative-care measures should be interpreted cautiously.
Introduction
Chronic obstructive pulmonary disease (COPD) is a progressive, life-limiting illness associated with substantial symptom burden, repeated exacerbations, functional decline, multimorbidity, and high healthcare utilization, particularly in advanced disease and near the end of life.1–6 Patients with advanced COPD commonly experience refractory dyspnea, fatigue, anxiety, depression, impaired quality of life, and recurrent hospitalizations, often alongside escalating care needs and caregiver burden.1–3 Despite this burden, palliative and supportive care integration in COPD has historically been less consistent than in cancer care, even when suffering and care needs may be comparable.1,4
A central challenge in advanced COPD is its clinical trajectory. Unlike many malignancies, COPD often follows a prolonged and unpredictable course marked by gradual decline, acute exacerbations, and partial recoveries. This pattern complicates prognostication, delays recognition of the terminal phase, and may postpone referral to palliative care and advance care planning.1,2,7,8 Qualitative research has similarly shown that patients and families often live with ongoing uncertainty regarding deterioration and death, which may contribute to reactive rather than anticipatory end-of-life care.9
The literature consistently shows that patients with COPD receive less palliative care and more intensive end-of-life treatment than patients with lung cancer. Comparative studies have shown that COPD decedents are more likely to undergo intensive care unit admission, prolonged hospital care, and invasive interventions, and are less likely to receive inpatient, outpatient, or home-based palliative care despite substantial symptom burden.1,3,4,10 At the same time, COPD is increasingly understood as a heterogeneous condition shaped not only by respiratory impairment but also by multimorbidity. Patients with advanced COPD frequently live with cardiovascular disease, renal disease, metabolic disorders, malignancy, neurocognitive decline, and psychiatric symptoms, all of which may influence symptom expression, exacerbation risk, treatment pathways, hospitalization, and survival.2,5,6,11,12 Low use of formal palliative care and other supportive strategies has been documented even among patients with advanced COPD who clearly carry substantial illness burden.2,8,13,14
An important dimension of this heterogeneity is sex. Growing evidence suggests that females and males with COPD differ in symptom burden, comorbidity patterns, and outcomes. Women with COPD have often been described as experiencing greater dyspnea burden and poorer health-related quality of life, whereas men more often carry a heavier burden of cardiovascular and renal comorbidity.11,12,15–17 Sex differences have also been reported in mortality, comorbidity distribution, and multidimensional disease severity.15–18 However, these two lines of literature, underuse of palliative care in COPD and sex-related differences in COPD burden, have rarely been integrated in studies focused specifically on end-of-life trajectories.1,3,4,8,10–12,14–18 Emerging work in advanced lung disease has further suggested that gender and sex disparities may persist toward the end of life, reinforcing the need to examine these patterns directly in advanced COPD populations.19
In Israel, this question is especially relevant. COPD care is delivered within a universal health insurance system, yet broad structural coverage does not guarantee optimal chronic respiratory care. The Israeli COPD country profile describes relatively strong access together with ongoing gaps in national strategy, rehabilitation capacity, workforce availability, co-payment burden, and consistent implementation of guideline-based care.20 Israeli data further suggest that COPD remains underdiagnosed and sub-optimally managed. Screening data among smokers aged 45 years and older showed substantial under-recognition of COPD.21 In addition, national research has documented high readmission and mortality after hospitalization for COPD exacerbations as well as deficiencies in discharge management, including low rates of prescription of long-acting bronchodilators, pulmonary rehabilitation referral, and smoking cessation counseling.22,23
Previous studies have described sex differences in COPD symptoms, comorbidity, and outcomes, but fewer have examined documented multimorbidity and service use specifically among decedents with severe airflow limitation and substantial symptom-management needs. Focusing on deceased patients permits description of records accumulated before death; however, it does not establish the prognosis of people currently living with advanced COPD and cannot reconstruct complete end-of-life trajectories. The principal contribution of this study is therefore a sex-stratified description of multimorbidity within a large Israeli health-system cohort, supplemented by selected administrative indicators of healthcare use and formally coded palliative care encounters.
The study aimed to compare women and men with respect to demographic characteristics, documented multimorbidity, selected healthcare-use indicators, formally coded palliative care encounters, and time from first documented COPD diagnosis to death. Because symptom scales, longitudinal spirometry, patient preferences, place of death, and complete hospital and supportive-care pathways were unavailable, the analysis was designed as a descriptive study of recorded indicators rather than a comprehensive evaluation of end-of-life care.
Methods
Aim
This study aimed to describe sex-based differences in demographic characteristics, multimorbidity burden, selected documented healthcare utilization indicators, documented palliative care encounters, and time from first documented COPD diagnosis to death among deceased patients with advanced COPD. The analysis was intended as a descriptive comparison of routinely recorded clinical and service-use indicators, rather than as a complete reconstruction of end-of-life healthcare utilization or care trajectories.
Study Design and Setting
This retrospective population-based cohort study used routinely collected administrative and clinical data from Clalit Health Services, the largest health maintenance organization in Israel. The study included deceased patients with advanced COPD during 2016–2020 and 2022–2023. The year 2021 was excluded a priori as the period designated in the study protocol to minimize distortion in healthcare utilization, respiratory morbidity, and mortality patterns; 2020 remained part of the prespecified observation window. The analysis was framed as a descriptive comparison by sex rather than as a complete reconstruction of end-of-life care trajectories, because the available dataset did not include full information on patient preferences, goals-of-care discussions, place of death, all specialist involvement, or detailed transitions between care settings.
Ethics Approval and Consent to Participate
The study was conducted in accordance with ethical standards for research using routinely collected health data. The study protocol was reviewed and approved by the Clalit Health Services Institutional Review Board Committee (approval number: MMC-0316-23). Because the study used retrospective, de-identified administrative and clinical data, the requirement for informed consent was waived by the approving ethics committee in accordance with institutional policy and applicable regulations.
Data Source
Data were extracted from Clalit Health Services electronic health records and administrative databases. The dataset included demographic characteristics, year of birth, sex, sector/ethnicity variables available in the administrative system, country of birth where available, year of COPD diagnosis, clinical diagnoses, pulmonary function indicators, Charlson Comorbidity Index, medication dispensations, selected laboratory and imaging records, inpatient and ambulatory service-use records, coded procedures, and documented palliative care service use. Variables were extracted at the patient level and analyzed using de-identified data.
Study Population
Eligible patients were aged 55 years or older at death and met the study’s operational definition of advanced COPD: a documented COPD diagnosis, post-bronchodilator FEV1/FVC <0.70, FEV1 <50% predicted, and at least two opioid prescriptions or dispensations during the final year of life. The spirometry thresholds indicate severe airflow limitation but do not, by themselves, measure symptom burden.
The opioid requirement was a cohort-selection criterion intended to identify patients with substantial symptom-management needs in the context of severe airflow limitation. It should not be interpreted as a validated marker of refractory dyspnea, because opioids may have been prescribed for pain, breathlessness, another comorbidity, or palliative symptom management. CAT (COPD Assessment Test) and mMRC (modified Medical Research Council) scores were unavailable. Consequently, the cohort represents a selected subgroup meeting this operational definition rather than the broader population of people living with advanced COPD. The opioid criterion may have preferentially selected patients with greater symptom burden or more frequent contact with supportive or palliative care services, thereby introducing potential selection bias and limiting the generalizability of the findings.24,25
Patients were excluded if they were younger than 55 years, had a concurrent diagnosis of asthma, had lung cancer, were listed for lung transplantation, or had admissions to post-acute care units or geriatric centers during the predefined observation period. Asthma was excluded to reduce diagnostic overlap. Lung cancer was excluded to reduce confounding from cancer-specific end-of-life pathways; therefore, the malignancy variable in this study refers to malignancies other than lung cancer. Patients admitted to post-acute care units or geriatric centers during the predefined observation period were excluded because these settings represent distinct care pathways characterized by facility-based rehabilitation, multidisciplinary management, and transitions between acute and institutional care. Healthcare utilization and palliative-care activity in these settings may be recorded through different administrative and clinical processes and were not captured with sufficient consistency in the available dataset to permit valid comparison with patients managed within the included inpatient and ambulatory pathways. Discharge to long-term or skilled nursing care is also associated with clinical complexity and subsequent healthcare utilization among patients with COPD. Accordingly, this predefined exclusion was intended to improve the comparability of the analyzed healthcare-use indicators rather than to imply that patients receiving post-acute or geriatric care were clinically unimportant.26,27
Diagnostic Coding
Administrative diagnoses were identified using ICD-9-CM diagnostic codes recorded in the Clalit Health Services database. COPD was identified using codes 490, 491, 492, and 496. Asthma was identified using code 493, and lung cancer using code 162.x. Comorbidities included in the Charlson Comorbidity Index and selected binary comorbidity variables were identified from documented diagnosis codes in the administrative database. Because all patients met inclusion criteria for advanced COPD, COPD was treated as the index condition for cohort entry and was not interpreted as an explanatory comorbidity in sex comparisons.
Variables and Measures
The main descriptive variables were sex, age at death, socioeconomic score where available, Charlson Comorbidity Index, total comorbidity count, selected comorbidities, selected healthcare utilization indicators, documented palliative care encounters, and time from first documented COPD diagnosis to death.
Comorbidity burden was summarized using the Charlson Comorbidity Index (CCI), a weighted index of specified comorbid conditions originally developed to estimate mortality risk.28 CCI scores were derived from diagnoses recorded in the Clalit administrative database. Selected coexisting conditions included congestive heart failure, myocardial infarction, peripheral vascular disease, cerebrovascular disease, dementia, renal disease, diabetes mellitus, liver disease, peptic ulcer disease, rheumatic/connective tissue disease, hemiplegia or paraplegia, lymphoma, leukemia, HIV/AIDS, and malignancy other than lung cancer.
The inhaler variable represented the number of documented COPD-related inhaler medication records available in the administrative dataset during the predefined observation period. It did not represent the number of inhalations taken per day, prescribed daily doses, or treatment adherence. The procedure variable represented the count of coded diagnostic, respiratory, imaging, laboratory, and selected service procedures available in the Clalit administrative extract during the predefined observation period. The source dataset contained inpatient and ambulatory service records; however, the procedure count did not provide a complete reconstruction of all procedures performed in every care setting. It was not a count of clinical visits and is therefore reported as a partial administrative indicator rather than as comprehensive healthcare utilization. Documented palliative care exposure was identified using dedicated palliative care encounter codes recorded in the Clalit administrative database. These included nurse-unit palliative encounters and physician-unit palliative encounters, according to the service-provider classification used in the administrative system. These variables reflect formally coded palliative care activity only and may underestimate supportive or palliative care delivered by primary care, pulmonary, geriatric, hospital, or community teams when such care was not coded under dedicated palliative care encounter categories. Because palliative care encounter counts were available only for patients with recorded palliative activity, analyses of nurse-unit and physician-unit palliative encounters were conducted as available-case analyses. Given the small numbers, these findings were interpreted as exploratory and descriptive rather than as definitive evidence of sex differences in palliative care access or intensity.
Time-to-event was defined as the interval between the first COPD diagnosis recorded in the Clalit database and death. The starting point is therefore first documentation within this health system, not necessarily biological disease onset or the date on which diagnostic criteria were first met. Late recognition and incomplete capture of diagnoses made before enrollment in Clalit may shorten the observed interval. Because all participants had died, this measure describes documentation-to-death time in a selected decedent cohort and is not an estimate of prognosis for all patients with COPD. For each variable, analyses used the number of patients with a recorded value (available-case analysis); denominators are reported in the text or tables when they differ from the full cohort. No missing values were imputed. Absence of a dedicated palliative-care code was not assumed to mean absence of all supportive or palliative care. The very small palliative-care subsamples were analyzed descriptively and are considered exploratory.
Statistical Analysis
Statistical analyses were performed using IBM SPSS Statistics for Windows, version 27.0 (IBM Corp., Armonk, NY, USA) and R, version 3.5 (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables were summarized as counts and percentages, and continuous variables as mean and standard deviation or median and interquartile range, as appropriate. Normality of continuous variables was assessed using the Kolmogorov–Smirnov and Shapiro–Wilk tests. Between-sex comparisons were performed using independent-samples t tests for approximately normally distributed continuous variables and Mann–Whitney U-tests for non-normally distributed continuous variables. Categorical variables were compared using chi-square or Fisher’s exact tests, as appropriate. Because the large cohort could yield statistically significant differences of limited clinical magnitude, p values were interpreted alongside absolute differences and effect sizes. Phi values of approximately 0.10 were regarded as small associations; findings with very small absolute cell counts were treated as exploratory. Multiple comorbidity comparisons were conducted without adjustment for multiplicity. Accordingly, statistically significant findings, particularly those involving small absolute differences or rare conditions, were considered exploratory and interpreted alongside their effect sizes.
Time from first documented COPD diagnosis to death was described using the Kaplan-Meier method and compared between women and men using the Log rank test. Median documentation-to-death time and corresponding 95% confidence intervals were reported. Because the cohort included only deceased patients, this analysis was descriptive and should not be interpreted as estimating prognosis or survival among all patients living with advanced COPD. No multivariable survival model was performed because the study’s primary objective was descriptive and the available covariates did not permit a clearly defined baseline-adjusted prognostic analysis.
Results
The cohort included 2261 deceased patients with advanced COPD, of whom 547 were female and 1714 were male. Table 1 presents cohort characteristics by sex. Mean CCI was statistically lower among women than men (8.39 vs 8.83; p=0.005), but the absolute difference was 0.44 points. Because no prespecified minimal clinically important difference exists for this comparison, the clinical importance of this small mean difference is uncertain. Inhaler treatment counts did not differ by sex, whereas men had a higher number of documented procedures. Median time from first documented COPD diagnosis to death was slightly longer among women than men, but the difference was not statistically significant.
Table 1.
Cohort Characteristics by Sex
| Characteristic | Female Patients n=547 | Male Patients n=1714 | Total Cohort N=2261 | p-value |
|---|---|---|---|---|
| Age at death, mean ± SD | 79.98 ± 9.54 | 77.04 ± 9.16 | 77.75 ± 9.34 | <0.001 |
| Charlson Comorbidity Index, mean ± SD | 8.39 ± 3.02 | 8.83 ± 3.20 | 8.73 ± 3.16 | 0.005 |
| Total comorbidity count, mean ± SD | 3.65 ± 1.70 | 4.18 ± 1.87 | 4.05 ± 1.84 | <0.001 |
| Number of documented inhaler medication records, mean ± SD | 8.35 ± 5.74 | 8.33 ± 5.73 | 8.33 ± 5.73 | 0.952 |
| Number of documented coded procedures, mean ± SD | 4.87 ± 5.01 | 6.48 ± 8.27 | 6.12 ± 7.70 | <0.001 |
| Time from first documented COPD diagnosis to death, median months | 33.00 | 29.00 | 30.00 | 0.328 |
Notes: Values are presented as mean ± SD unless otherwise stated. Percentages are calculated within sex groups where applicable. Documented coded procedures were available for 1706 patients: 376 female patients and 1330 male patients. Procedures refer to coded service and procedure records documented in the administrative database during the available observation period. Survival was calculated from documented COPD diagnosis to death. The p-value for survival refers to the Log rank test.
Women comprised 24.2% and men 75.8% of the cohort. Smoking exposure in pack-years was unavailable; therefore, the sex imbalance and observed comorbidity patterns cannot be attributed directly to smoking history. Possible explanations include historical sex differences in smoking and occupational exposure, differences in COPD recognition, and selection introduced by the opioid-based eligibility criterion, but these mechanisms could not be tested. All included patients met the spirometry eligibility thresholds of FEV1/FVC <0.70 and FEV1 <50% predicted. Sex-stratified continuous spirometry values were not available for the present analysis. Consequently, between-sex differences in FEV1 could not be assessed beyond confirming that all participants met the predefined eligibility thresholds (Table 1).
Sex differences were observed in several documented comorbidities. Female patients had a higher prevalence of dementia than male patients (13.9% vs 9.6%, p=0.005) and rheumatic/connective tissue disease (8.4% vs 3.5%, p<0.001). Female patients also had a higher recorded prevalence of HIV/AIDS (1.6% vs 0.4%, Fisher exact p=0.006), although absolute numbers were very small and this finding should be interpreted cautiously.
Male patients had a higher prevalence of myocardial infarction (37.6% vs 21.4%, p<0.001), renal disease (44.9% vs 31.3%, p<0.001), peripheral vascular disease (32.4% vs 20.7%, p<0.001), and diabetes mellitus (52.2% vs 47.0%, p=0.033). No statistically significant sex differences were observed for cerebrovascular disease, congestive heart failure, hemiplegia/paraplegia, leukemia, liver disease, lymphoma, malignancy other than lung cancer, or peptic ulcer disease (Table 2).
Table 2.
Sex Differences in Comorbidities
| Condition | Female n (%) | Male n (%) | Total n (%) | p-value | Phi |
|---|---|---|---|---|---|
| AIDS/HIV | 9 (1.6) | 7 (0.4) | 16 (0.7) | 0.006 | −0.063 |
| Cerebrovascular disease | 72 (13.2) | 264 (15.4) | 336 (14.9) | 0.200 | 0.027 |
| Congestive heart failure | 268 (49.0) | 896 (52.3) | 1164 (51.5) | 0.181 | 0.028 |
| Dementia | 76 (13.9) | 165 (9.6) | 241 (10.7) | 0.005 | −0.059 |
| Diabetes mellitus | 257 (47.0) | 895 (52.2) | 1152 (50.9) | 0.033 | 0.045 |
| Hemiplegia/paraplegia | 21 (3.8) | 68 (4.0) | 89 (3.9) | 0.893 | 0.003 |
| Leukemia | 9 (1.6) | 45 (2.6) | 54 (2.4) | 0.191 | 0.027 |
| Liver disease | 45 (8.2) | 188 (11.0) | 233 (10.3) | 0.066 | 0.039 |
| Lymphoma | 23 (4.2) | 63 (3.7) | 86 (3.8) | 0.573 | −0.012 |
| Malignancy other than lung cancer | 177 (32.4) | 609 (35.5) | 786 (34.8) | 0.175 | 0.029 |
| Myocardial infarction | 117 (21.4) | 644 (37.6) | 761 (33.7) | <0.001 | 0.147 |
| Peptic ulcer disease | 97 (17.7) | 334 (19.5) | 431 (19.1) | 0.363 | 0.019 |
| Peripheral vascular disease | 113 (20.7) | 556 (32.4) | 669 (29.6) | <0.001 | 0.111 |
| Renal disease | 171 (31.3) | 769 (44.9) | 940 (41.6) | <0.001 | 0.118 |
| Rheumatic/connective tissue disease | 46 (8.4) | 60 (3.5) | 106 (4.7) | <0.001 | −0.099 |
Notes: Percentages are calculated within sex groups: women n=547; men n=1714. P values were calculated using Pearson chi-square tests except for HIV/AIDS, for which Fisher’s exact test was used. Phi coefficients are based on the corresponding 2×2 tables; negative values indicate higher prevalence among women and positive values indicate higher prevalence among men.
Although several comparisons reached statistical significance, all phi coefficients were small (absolute range among significant comparisons, 0.045–0.147). These findings therefore indicate modest associations and should not be interpreted as large sex effects. The HIV/AIDS comparison involved only 16 patients and is exploratory (Table 2).
Selected documented utilization indicators are presented in Table 3. The number of documented inhaler medication records did not differ between women and men (8.35±5.74 vs 8.33±5.73, p=0.952). Men had a higher mean number of documented coded procedures than women (6.48±8.27 vs 4.87±5.01, p<0.001), although the standardized difference was small (Cohen’s d=0.21). Because procedure data were incomplete with respect to setting and did not capture all hospital activity, this finding should be interpreted only as a difference in the coded records available for analysis, not as evidence that men received more total healthcare or more intensive end-of-life care (Table 3).
Table 3.
COPD-Related Treatment Indicators by Sex
| Outcome | Women n | Women mean ± SD | Men n | Men mean ± SD | Test | p-value |
|---|---|---|---|---|---|---|
| Documented inhaler medication records | 547 | 8.35 ± 5.74 | 1714 | 8.33 ± 5.73 | t = 0.061 | 0.952 |
| Documented coded procedures | 376 | 4.87 ± 5.01 | 1330 | 6.48 ± 8.27 | t = −4.670 | <0.001 |
Dedicated palliative-care encounter codes were present for only 27 patients in each provider-specific available-case analysis. These counts do not represent the prevalence of all palliative or supportive care in the cohort. Given the extremely small and selected denominators, between-sex tests are unstable and potentially misleading; Table 4 therefore presents counts descriptively without inferential claims (Table 4).
Table 4.
Descriptive Summary of Formally Coded Palliative-Care Encounters Among Patients with Recorded Activity
| Variable | Female n | Female Mean ± SD | Male n | Male Mean ± SD |
|---|---|---|---|---|
| Nurse-unit palliative visits | 4 | 2.00 ± 2.00 | 23 | 1.52 ± 0.85 |
| Physician-unit palliative visits | 6 | 1.00 ± 0.00 | 21 | 1.86 ± 0.85 |
Notes: Values are based only on patients with at least one recorded encounter in the relevant provider category and do not represent the prevalence of palliative care in the full cohort. Because of the very small available-case subsamples, no inferential comparisons by sex were performed.
Kaplan-Meier analysis included 2261 deceased patients. Median time from first documented COPD diagnosis to death was 33.0 months among women (95% CI, 29.42–36.58) and 29.0 months among men (95% CI, 27.08–30.92). The survival distributions did not differ significantly by sex (log-rank χ2=0.955, df=1, p=0.328). These estimates describe documentation-to-death intervals within this selected decedent cohort and do not represent general COPD prognosis (Table 5).
Table 5.
Kaplan-Meier Estimates of Time from First Documented COPD Diagnosis to Death by Sex
| Sex | N | Events, n (%) | Median Documentation-to-Death Time, Months (SE) | 95% CI | Log-Rank p-value |
|---|---|---|---|---|---|
| Female | 547 | 547 (100.0) | 33.00 (1.83) | 29.42–36.58 | 0.328 |
| Male | 1714 | 1714 (100.0) | 29.00 (0.98) | 27.08–30.92 | 0.328 |
| Overall | 2261 | 2261 (100.0) | 30.00 (0.86) | 28.31–31.69 | - |
The observed medians indicate a short interval between first recorded COPD diagnosis and death, but they do not establish that biological disease onset occurred late. COPD may have been present before its first coded diagnosis or before entry into the available records. These estimates are consequently interpreted as documentation-to-death intervals, not disease-duration or population-survival estimates.
Discussion
The principal finding of this study was that women and men in a selected cohort of deceased patients meeting an operational definition of advanced COPD had different documented multimorbidity profiles. Men more frequently had myocardial infarction, renal disease, peripheral vascular disease, and diabetes mellitus, whereas women more frequently had dementia and rheumatic or connective tissue disease. Although several differences were statistically significant, the absolute difference in mean Charlson Comorbidity Index (CCI) was small and the phi coefficients indicated weak associations. Men also had more coded procedures, but the effect size was small and the available procedure data did not represent complete healthcare utilization. Accordingly, sex-related differences in documented multimorbidity constitute the strongest contribution of this analysis, whereas findings concerning healthcare use, palliative care, and documentation-to-death time are more limited and should be interpreted cautiously.
The greater frequency of cardiometabolic, vascular, and renal conditions among men is consistent with previous studies describing sex-related variation in COPD multimorbidity.12,15–17 These coexisting conditions may complicate respiratory management and increase the need for coordination across pulmonary, cardiovascular, renal, and primary care services. Conversely, the higher prevalence of dementia among women underscores the potential relevance of cognitive assessment, caregiver involvement, communication support, and advance care planning.8,15,16 These implications should not be interpreted as evidence that sex caused the observed differences. Sex functioned as a descriptive stratification variable, and the associations may also reflect differences in age, smoking and occupational exposures, diagnosis, survival, or selection into the cohort. Because smoking exposure in pack-years and occupational exposure data were unavailable, these possible explanations could not be examined directly.
The findings concerning documented procedures require similarly cautious interpretation. Although men had a higher mean number of coded procedures, the standardized difference was small. Moreover, the administrative extract did not provide a complete account of procedures across all care settings or during specific intervals before death. The procedure variable should therefore be understood as a partial indicator of recorded service use rather than as evidence that men received more intensive healthcare or more burdensome end-of-life treatment. More complete analyses would require linked information on hospital admissions, emergency department visits, intensive care, specialist involvement, procedures performed across settings, and transitions between care environments.
Formally coded palliative-care encounters were uncommon and available only for very small subsamples. These codes captured encounters assigned to dedicated palliative-care services but did not identify palliative approaches delivered by pulmonary, primary-care, geriatric, hospital, or community teams. The findings therefore cannot establish the prevalence, timing, quality, or intensity of palliative care, nor can they support firm conclusions regarding differences between women and men. They should instead be regarded as preliminary observations concerning formal coding within the available data. This distinction is important because previous research has shown that patients with COPD may receive limited specialist palliative care despite substantial illness burden.1,3,4,8,10,14 Future studies should incorporate referral timing, care setting, symptom severity, goals-of-care documentation, advance care planning, place of death, and palliative care delivered outside formally designated services.
Median time from first documented COPD diagnosis to death was 33 months among women and 29 months among men, and the Kaplan-Meier distributions did not differ significantly. These estimates do not necessarily represent the duration of COPD because the first diagnosis recorded in the database may have occurred after symptom onset or after diagnostic criteria were initially met. Incomplete capture of diagnoses recorded before entry into the available system may also have shortened the observed interval. Furthermore, because the cohort included only deceased patients, the analysis describes documentation-to-death time within this selected cohort and should not be interpreted as prognosis for the broader population of people living with advanced COPD.
The operational definition of advanced COPD also shapes interpretation of the findings. All participants met predefined spirometry thresholds, but sex-stratified continuous spirometry indices were unavailable for analysis. CAT and mMRC scores were also unavailable; consequently, respiratory symptom burden could not be assessed using validated instruments. In addition, the requirement for at least two opioid prescriptions or dispensations during the final year of life may have selected patients with greater symptom burden, pain from comorbid conditions, or more frequent contact with supportive or palliative services. Opioid use should therefore not be interpreted as a validated proxy for refractory dyspnea or as evidence that palliative care was provided.24,25 The findings apply to patients meeting this operational definition and should not be generalized to all people with advanced COPD.
From a clinical perspective, the results support assessment of the entire multimorbidity profile rather than management of COPD as an isolated respiratory disorder. Cardiometabolic and renal disease may warrant particular attention among men in similar populations, while cognitive impairment and caregiver-related needs may require greater attention among women. These are group-level patterns, not deterministic clinical rules, and individual assessment should remain guided by age, function, frailty, symptoms, comorbidities, personal preferences, and clinical trajectory. Within the Israeli health system, systematic identification of multimorbidity may help inform coordinated follow-up and supportive-care planning; however, the present data do not establish which service model would improve outcomes.
Strengths and Limitations
This study used a large population-based dataset from Israel’s largest health maintenance organization and examined sex-related variation across demographic characteristics, documented multimorbidity, selected healthcare-use indicators, formally coded palliative-care encounters, and documentation-to-death time. Its focus on deceased patients meeting predefined clinical and administrative criteria enabled examination of records accumulated before death in a population with severe COPD. Nevertheless, this design and the available data impose important limitations.
First, the retrospective analysis depended on diagnoses, prescriptions, procedures, and service codes recorded in structured administrative and clinical data. It did not capture unrecorded symptoms, informal supportive care, goals-of-care discussions, patient preferences, caregiver involvement, or communication quality. No data were available on smoking pack-years, occupational exposures, frailty, functional status, longitudinal symptom burden, CAT or mMRC scores, or sex-stratified continuous spirometry values. These omissions restrict interpretation of the unequal sex distribution and the observed multimorbidity patterns.
Second, the operational definition required severe airflow limitation and at least two opioid prescriptions or dispensations during the final year of life. Because opioids may have been prescribed for dyspnea, pain, other comorbidities, or palliative symptom management, this criterion may have introduced selection bias and does not provide a validated measure of respiratory symptom burden.24,25 The results therefore apply only to the selected subgroup meeting the study definition.
Third, exclusion of patients admitted to post-acute care units or geriatric centers improved comparability of the available administrative indicators but may have excluded patients with greater frailty, functional impairment, cognitive impairment, rehabilitation needs, or care-transition complexity. This exclusion may have led to underestimation of healthcare utilization and palliative-care needs and limits generalizability to the most functionally impaired patients with advanced COPD.26,27,29
Fourth, the healthcare-use variables were incomplete. The inhaler measure reflected documented medication records and did not represent daily inhalations, adherence, or treatment effectiveness. Procedure counts captured only coded records available in the extract and did not provide a comprehensive account of inpatient and ambulatory procedures, hospitalizations, emergency care, intensive care, specialist involvement, place of death, or transitions between settings. Similarly, dedicated palliative-care codes may have underestimated palliative approaches delivered outside formally classified services.
Finally, the predefined exclusion of 2021 resulted in a discontinuous observation timeframe, and residual calendar-period effects cannot be excluded. The Kaplan-Meier analysis was also limited by the decedent-only design and by use of the first documented COPD diagnosis as the starting point. Diagnostic delay and incomplete historical capture may have shortened the observed documentation-to-death interval; therefore, these estimates should not be interpreted as disease duration or general COPD prognosis. Multiple comorbidity comparisons were performed without adjustment for multiplicity; consequently, nominally significant findings, especially those involving rare conditions, should be regarded as exploratory.
Conclusions
Among deceased patients meeting this study’s operational definition of advanced COPD, women and men differed primarily in documented multimorbidity profiles. Men more often had cardiometabolic, vascular, and renal diagnoses, whereas women more often had dementia and rheumatic/connective tissue disease; however, association sizes were small. Differences in coded procedures were also small and based on partial administrative data. Formally coded palliative-care encounters were too uncommon for firm sex comparisons, and documentation-to-death intervals did not differ significantly. These findings support attention to multimorbidity in advanced COPD while underscoring the need for future prospective studies incorporating smoking exposure, symptom scales, serial spirometry, frailty, patient preferences, and complete cross-setting healthcare and palliative-care data.
Acknowledgments
The authors thank Clalit Health Services for enabling use of approved de-identified administrative and clinical data within the institutional research governance framework.
Funding Statement
The authors received no specific funding for this work.
Abbreviations
CAT, COPD Assessment Test; CCI, Charlson Comorbidity Index; CI, Confidence Interval; COPD, Chronic Obstructive Pulmonary Disease; FEV1, Forced Expiratory Volume in 1 Second; FVC, Forced Vital Capacity; HIV, Human Immunodeficiency Virus; ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification; IRB, Institutional Review Board; mMRC, Modified Medical Research Council; SE, Standard Error.
Data Sharing Statement
The datasets generated and/or analyzed during the current study are not publicly available due to institutional and privacy restrictions but are available from the corresponding author on reasonable request and with permission of Clalit Health Services.
Ethics Approval and Consent to Participate
The study was conducted in accordance with the Declaration of Helsinki. The protocol was reviewed and approved by the Clalit Health Services Institutional Review Board Committee (approval number MMC-0316-23). Because the study used retrospective, de-identified administrative and clinical data, the requirement for individual informed consent was waived by the approving ethics committee in accordance with institutional policy and applicable regulations.
Author Contributions
All authors made a significant contribution to the work reported, whether in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; agreed on the journal to which the article was submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors declare that they have no conflicts of interest relevant to this work.
References
- 1.Beernaert K, Cohen J, Deliens L, et al. Referral to palliative care in COPD and other chronic diseases: a population-based study. Respir Med. 2013;107(11):1731–12. doi: 10.1016/j.rmed.2013.06.003 [DOI] [PubMed] [Google Scholar]
- 2.Gershon AS, Maclagan LC, Luo J, et al. End-of-life strategies among patients with advanced chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2018;198(11):1389–1396. doi: 10.1164/rccm.201803-0592OC [DOI] [PubMed] [Google Scholar]
- 3.Suen AO, Bischoff K, Iyer AS, et al. Differences in health care and palliative care use at the end of life: a comparison study among lung cancer, COPD, and idiopathic pulmonary fibrosis. Chest. 2024;166(6):1487–1496. doi: 10.1016/j.chest.2024.06.3807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Butler SJ, Ellerton L, Gershon AS, Goldstein RS, Brooks D. Comparison of end-of-life care in people with chronic obstructive pulmonary disease or lung cancer: a systematic review. Palliat Med. 2020;34(8):1030–1043. doi: 10.1177/0269216320929556 [DOI] [PubMed] [Google Scholar]
- 5.Divo M, Cote C, de Torres JP, et al. Comorbidities and risk of mortality in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2012;186(2):155–161. doi: 10.1164/rccm.201201-0034OC [DOI] [PubMed] [Google Scholar]
- 6.Vanfleteren LEGW, Spruit MA, Groenen M, et al. Clusters of comorbidities based on validated objective measurements and systemic inflammation in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2013;187(7):728–735. doi: 10.1164/rccm.201209-1665OC [DOI] [PubMed] [Google Scholar]
- 7.Tsai SH, Hung JY, Su PF, et al. Chronic obstructive pulmonary disease trajectory: severe exacerbations and dynamic change in health-related quality of life. BMJ Open Respir Res. 2024;11:e002037. doi: 10.1136/bmjresp-2023-002037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Janssen DJA, Bajwah S, Boon MH, et al. European Respiratory Society clinical practice guideline: palliative care for people with chronic obstructive pulmonary disease or interstitial lung disease. Eur Respir J. 2023;62(2):2202014. doi: 10.1183/13993003.02014-2022 [DOI] [PubMed] [Google Scholar]
- 9.Ek K, Andershed B, Sahlberg-Blom E, Ternestedt BM. The unpredictable death: the last year of life for patients with advanced COPD: relatives’ stories. Palliat Support Care. 2015;13(5):1213–1222. doi: 10.1017/S1478951514001151 [DOI] [PubMed] [Google Scholar]
- 10.Kuo LC, Chen JH, Lee CH, Tsai CW, Lin CC. End-of-life health care utilization between chronic obstructive pulmonary disease and lung cancer patients. J Pain Symptom Manage. 2019;57(5):933–943. doi: 10.1016/j.jpainsymman.2019.01.011 [DOI] [PubMed] [Google Scholar]
- 11.Buttery SC, Zysman M, Vikjord SAA, Hopkinson NS, Jenkins C, Vanfleteren LEGW. Contemporary perspectives in COPD: patient burden, the role of gender and trajectories of multimorbidity. Respirology. 2021;26(5):419–441. doi: 10.1111/resp.14032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen W, Thomas J, Sadatsafavi M, FitzGerald JM. Risk of cardiovascular comorbidity in patients with chronic obstructive pulmonary disease: a systematic review and meta-analysis. Lancet Respir Med. 2015;3(8):631–639. doi: 10.1016/S2213-2600(15)00241-6 [DOI] [PubMed] [Google Scholar]
- 13.Henoch I, Ekberg-Jansson A, Löfdahl CG, Strang P. Benefits, for patients with late-stage chronic obstructive pulmonary disease, of being cared for in specialized palliative care compared to hospital: a nationwide register study. BMC Palliat Care. 2017;16:13. doi: 10.1186/s12904-017-0178-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Iyer AS, Goodrich CA, Dransfield MT, et al. End-of-life spending and healthcare utilization among older adults with chronic obstructive pulmonary disease. Am J Med. 2020;133(7):817–824.e1. doi: 10.1016/j.amjmed.2019.11.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lisspers K, Larsson K, Janson C, et al. Gender differences among Swedish COPD patients: results from the Arctic, a real-world retrospective cohort study. NPJ Prim Care Respir Med. 2019;29(1):45. doi: 10.1038/s41533-019-0157-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Li N, Li X, Chen S, et al. Sex differences in comorbidities and mortality risk among patients with chronic obstructive pulmonary disease: a study based on NHANES data. BMC Pulm Med. 2023;23(1):481. doi: 10.1186/s12890-023-02771-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ekström MP, Jogréus C, Ström KE. Comorbidity and sex-related differences in mortality in oxygen-dependent chronic obstructive pulmonary disease. PLoS One. 2012;7(4):e35806. doi: 10.1371/journal.pone.0035806 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.de Torres JP, Cote CG, López MV, et al. Sex differences in mortality in patients with COPD. Eur Respir J. 2009;33(3):528–535. doi: 10.1183/09031936.00096108 [DOI] [PubMed] [Google Scholar]
- 19.Chen X, Goh N, Dunn S, Smallwood N. Gender disparities in advanced lung diseases: do they persist towards the end of life? Am J Hosp Palliat Care. 2025;42(11):1128–1135. doi: 10.1177/10499091241299776 [DOI] [PubMed] [Google Scholar]
- 20.Global Allergy & Airways Patient Platform. Israel: COPD country profile. Respiratory Health. 2025. Available from: https://www.respiratoryhealth.org/copd-country/israel. Accessed August 1, 2026.
- 21.Stav D, Raz M. Prevalence of chronic obstructive pulmonary disease among smokers aged 45 and up in Israel. Isr Med Assoc J. 2007;9(11):800–802. [PubMed] [Google Scholar]
- 22.Shani M, Comaneshter D, Segel MJ, et al. The importance of having good quality indicators for care of patients with COPD: a look at hospital readmission rates. Isr J Health Policy Res. 2022;11:17. doi: 10.1186/s13584-022-00528-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bar-Shai A, Freund O, Ovdat T, Segel MJ, Klempfner R, Elis A. Management of acute COPD exacerbations in the internal medicine departments in Israel: a national survey. Front Med. 2023;10:1174148. doi: 10.3389/fmed.2023.1174148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ekström M, Nilsson F, Abernethy AA, Currow DC. Effects of opioids on breathlessness and exercise capacity in chronic obstructive pulmonary disease: a systematic review. Ann Am Thorac Soc. 2015;12(7):1079–1092. doi: 10.1513/AnnalsATS.201501-034OC [DOI] [PubMed] [Google Scholar]
- 25.Barnes H, McDonald J, Smallwood N, Manser R. Opioids for the palliation of refractory breathlessness in adults with advanced disease and terminal illness. Cochrane Database Syst Rev. 2016;3(3):CD011008. doi: 10.1002/14651858.CD011008.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Chow R, So OW, JHB I, et al. Predictors of readmission for patients with chronic obstructive pulmonary disease: a systematic review. Int J Chron Obstruct Pulmon Dis. 2023;18:2581–2617. doi: 10.2147/COPD.S418295 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hsiao YL, Bass EB, Wu AW, et al. Preventing avoidable rehospitalizations through standardizing management of chronic conditions in skilled nursing facilities. J Am Med Dir Assoc. 2023;24(12):1910–1917.e3. doi: 10.1016/j.jamda.2023.08.010 [DOI] [PubMed] [Google Scholar]
- 28.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–383. doi: 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
- 29.Ouslander JG, Diaz S, Hain D, Tappen R. Frequency and diagnoses associated with 7- and 30-day readmission of skilled nursing facility patients to a nonteaching community hospital. J Am Med Dir Assoc. 2011;12(3):195–203. doi: 10.1016/j.jamda.2010.02.015 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due to institutional and privacy restrictions but are available from the corresponding author on reasonable request and with permission of Clalit Health Services.