Discharge status and post-discharge healthcare costs after skeletal-related event hospitalizations among medicare patients with bone metastatic solid tumors or multiple myeloma.

Introduction

About 330,000 US adults developed bone metastases from solid tumors in 2012 [1] . These patients are at high risk for painful and costly bone complications, termed skeletal-related events (SREs), which include bone fractures, spinal cord compression, and radiation or surgery to the bone [2] , [3] , [4] . Patients with multiple myeloma (MM) are similarly at risk of SREs as a result of destructive bone lesions that are a defining characteristic of the disease [5] . >32,000 individuals were diagnosed with MM in 2019, with an estimated prevalence of>130,000 [6] .

Previous studies using commercial claims databases show that healthcare resource utilization and economic burden are increased in adults with bone metastases [7] , [8] , [9] or MM [10] , [11] who experience SREs. A study using the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database also showed increased healthcare resource utilization and Medicare costs associated with incident SREs in men with prostate cancer metastatic to bone [12] One study of 569 Danish women with SREs secondary to breast cancer and bone metastases showed that patients with multiple SREs experienced more visits and longer hospitalization stays during the 90 days after an SRE than patients with a single SRE [13]

Previous studies have shown that hospitalization cost is the primary contributor to SRE cost [7] , [8] , [9] , [10] , [11] . Resource use and associated costs are likely to be substantial after SRE hospitalizations, depending on hospital discharge status. Post-SRE hospitalization management and care is important, especially for older patients who require more comprehensive care. Medicare generally requires hospitals to engage in discharge planning to determine discharge locations for patients; however, because studies of post-discharge status after an SRE-related hospitalization are rare, data for use in such planning are lacking. A study of post-discharge status after total knee arthroplasty using Medicare data showed that 21% of patients were discharged to home, 38% to home health care, 31% to extended-care facilities, and 10% to rehabilitation [14] . In the general Medicare population, among>137 million hospital discharges, 26.3% were discharged to skilled nursing facilities (SNFs) or inpatient rehabilitation facilities [15] . In the current analysis, we aimed to understand discharge status and related post-discharge costs subsequent to an index SRE-related hospitalization among two cohorts of Medicare beneficiaries frequently affected by SREs: bone metastatic solid tumor patients and MM patients.

Methods

Study population and data Source

The study population included elderly fee-for-service (FFS) beneficiaries in the 2011–2015 Medicare 20% random sample. Medicare data include enrollment information, demographic characteristics, and medical claims from Part A, Part B, and Part D. The Medicare claims files contain information collected by Medicare to allow payment for healthcare services provided to Medicare beneficiaries in the US and its territories. Standard analytic files (SAFs) generated by the Centers for Medicare & Medicaid Services were used in this study. SAFs are available for each institutional claim type including inpatient, outpatient, SNF, home health agency (HHA), carrier, and hospice. Non-institutional Part B physician/supplier SAFs include final action claims for physician services, laboratory services (not laboratory values), and durable medical equipment (DME). Part D data include the prescription drug event file, which contains the National Drug Code for each drug, prescription dosing information, and drug costs. The 2011–2015 long-term care (LTC) Minimum Data Set (MDS) was used to identify LTC facility stays. The MDS is a standardized, primary screening and assessment tool for health status, which forms the foundation of the comprehensive assessment for all residents (regardless of payer) of LTC facilities certified to participate in Medicare or Medicaid.

Study cohorts and study design

This study included two study cohorts: (1) patients with bone metastases from solid tumors and (2) patients with MM. Patients with bone metastases were identified by at least one inpatient claim or two outpatient claims on different days within a 12-month interval carrying the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis code for secondary malignant neoplasm of bone or bone marrow (198.5). The bone metastasis date was defined as the earliest of (1) discharge date for inpatient claims and (2) second outpatient date. This algorithm is based on two previously published studies aiming to estimate prevalence of bone metastases using a Medicare FFS sample and the Truven Marketscan and PharMetrics Plus commercial claims databases [1] , [16] . Solid tumor types included prostate, breast (female only), lung, and other cancers during a 12-month interval before bone metastases were identified. ICD-9-CM diagnosis codes for specific cancer types appear in Appendix A . Cancer types were defined by at least one inpatient claim or two outpatient claims on different days in the 12-month interval. The following inclusion criteria were applied: (1) evidence of bone metastases from a solid tumor and subsequent SRE-related hospitalization (i.e., index SRE hospitalization) with discharge dates (i.e., index dates) between January 1, 2011, and September 30, 2015, (2) age ≥ 66 years on the index date, and (3) continuously enrolled in Medicare Parts A and B for ≥ 12 months before and on the index date. The following exclusion criteria were applied: (1) evidence of two or more solid tumors, (2) death on the index date, and (3) enrollment in a Medicare Advantage plan anytime during the 12 months before or on the index date.

Similarly, we defined patients with MM based on a validated algorithm requiring a combination of ICD-9-CM diagnosis codes ( Appendix B ) and diagnostic tests or treatment [17] . The following inclusion criteria were applied: (1) evidence of MM between January 1, 2011, and June 30, 2015; (2) evidence of subsequent SRE-related hospitalization (i.e., index SRE hospitalization) with discharge date (i.e., index dates) between January 1, 2011, and September 30, 2015; (3) age ≥ 66 years on the index date; and (4) continuously enrolled in Medicare Parts A and B for ≥ 12 months before and on the index date. The following exclusion criteria were applied: (1) death on the index date and (2) enrollment in a Medicare Advantage plan during the study period.

The detailed study design is displayed in Fig. 1 . For both study cohorts, the baseline period was 1 year prior to the index date. During the baseline period, patient demographics, Medicare/Medicaid dual eligible status, pre-index SRE-related hospitalizations (including the index SRE hospitalizations), and Charlson comorbidity index (CCI) were defined. The follow-up period was from the index date to the earliest of date of death, end of Medicare enrollment in both Part A and B, enrollment in a Medicare Advantage program, or December 31, 2015. During the follow-up period, we identified post-discharge facilities after index SRE hospitalizations and assessed lengths of stay and Medicare costs during the stays. We also assessed re-hospitalization within and after 30 days. Fig. 1 Study design.

Defining SREs

We used a claims-based algorithm to define SREs. Different methods have been used to identify SREs using claims in previous studies. In a recent study, Aly et al. compared different methods of identifying SREs from claims using Medicare-SEER data [18] . The current study adopted the method developed by Aly et al. and defined SREs with a 21-day window. SRE types included: spinal cord compression, fracture, and surgery or radiation to the bone. Codes and algorithms appear in Appendix C .

Defining discharge status from the index SRE hospitalization

As required by Medicare, discharging hospitals report patient discharge status using Discharge Disposition codes (e.g., code 03 identifies a SNF). However, in our exploratory analysis we found that the admitting facility subsequent to the index SRE hospitalization was not always the same as was indicated by the Discharge Disposition code provided by the discharging hospital, possibly due to claim error [19] . Therefore, we used the actual admitting facilities subsequent to index SRE hospitalizations to define discharge status.

Discharge status after the index SRE hospitalization was defined as admission to: SNF, rehabilitation facility, hospice, HHA, LTC hospital, LTC nursing home, re-hospitalization for any reason within 30 days, or re-hospitalization after 30 days. Except for LTC nursing homes, facilities were identified from Medicare Part A institutional claims and admissions were mutually exclusive. LTC nursing homes were identified from MDS data. Some patients were admitted to an LTC nursing home directly from the index SRE hospitalization. Only a few patients (less than 1%) were admitted directly to home without being admitted to any facility (including re-hospitalization) during follow-up.

Defining Medicare costs in an admitting facility after discharge from the index SRE hospitalization

The definition of Medicare costs was the same for all facilities except LTC nursing homes. Using SNFs as an example: When a patient was admitted to a SNF, Medicare costs included those for the SNF itself, outpatient services, physician visits/DME, and pharmacy costs defined from Medicare Part D claims. We included both Medicare paid amounts and patient or third party paid amounts, i.e., Medicare allowable. We calculated total Medicare allowable costs in a SNF and mean cost per SNF, dividing the total cost in the SNF by its total number of patients. Costs were adjusted to the 2015 US dollar.

Costs for LTC nursing homes were estimated by multiplying the length of LTC nursing home stays in days by the 2015 national daily nursing home cost of $235 per day [20] . We included other Medicare-covered costs such as outpatient services, physician visits/DME, and pharmacy during the LTC nursing home stay in the total cost.

Statistical analyses

All analyses were descriptive. For patient characteristics, we report number and percentage for categorical variables and mean (standard deviation [SD]) for continuous variables. For discharge status, we report number, percentage, and length of stay for each facility. The length of stay is reported as mean, SD, median, and interquartile range. For Medicare costs, we report mean, median, and SD cost for each facility.

Results

Patient characteristics

We identified 7988 (not including 405 deaths at discharge) patients with bone metastases from solid tumors and a subsequent SRE-related hospitalization. Separately, we identified 4277 (not including 170 deaths at discharge) patients with MM and a subsequent SRE-related hospitalization. Sample selection for both study cohorts is shown in Fig. 2 with inclusion and exclusion criteria. Baseline characteristics for both study cohorts are presented in Table 1 . Fig. 2 Sample selection and sample size. Table 1 Baseline Characteristics. Patients Bone Metastatic Solid Tumor Multiple Myeloma Characteristics n % n % n 7988 4277 Age; mean (SD) 76.9 (7.3) 76.6 (7.1) Age category, yrs. 66–69 1501 18.8 826 19.3 70–74 1976 24.7 1089 25.5 75–79 1665 20.8 856 20.0 ≥ 80 2846 35.6 1506 35.2 Sex Male 4026 50.4 1895 44.3 Female 3962 49.6 2382 55.7 Race White 7026 88.0 3530 82.5 Black 647 8.1 570 13.3 Other 315 3.9 177 4.1 Medicare-Medicaid dual eligible No 6966 87.2 3708 86.7 Yes 1022 12.8 569 13.3 Geographic region Northeast 1835 23.0 968 22.6 Midwest 2008 25.1 1029 24.1 South 2912 36.5 1602 37.5 West 1222 15.3 669 15.6 Missing 11 0.1 * * Pre-index SRE-related hospitalization LOS, days; mean (SD) †‡ 8.4 (9.1) 10.6 (10.4) Pre-index SRE-related hospitalization LOS categories, days †‡ 1–3 1661 20.8 803 18.8 4–7 3109 38.9 1392 32.5 ≥ 8 3218 40.3 2082 48.7 CCI categories † 0–4 118 1.5 2019 47.2 5–8 2638 33.0 1269 29.7 ≥ 9 5232 65.5 989 23.1 Tumor type § Breast cancer 1582 19.8 – – Prostate cancer 1743 21.8 – – Lung cancer 2393 30.0 – – Other cancer 2270 28.4 – – ‡ Including index SRE hospitalization. CCI, Charlson comorbidity index; LOS, length of stay; SD, standard deviation; SRE, skeletal related event. * Cells with n less than 11 suppressed. † Defined during the baseline period, i.e., one year prior to SRE index date which was the discharge date of the SRE-related hospitalization. § Defined during the 1 year prior to the bone metastases index date.

For bone metastases patients, mean (SD) age was 76.9 (7.3) years; 49.6% were women, 88.0% white, and 8.1% black. Mean (SD) length of pre-index SRE-related hospitalization (including the index SRE hospitalization) at baseline was 8.4 (9.1) days and mean (SD) CCI was 9.4 (1.9). Among the cancer types defined during the 1 year prior to the bone metastases diagnosis date, 19.8% had breast cancer, 21.8% prostate cancer, 30.0%, lung cancer, and 28.4% other cancers. As shown in Table 2 , the mean (SD) length of the index SRE hospitalization was 8.2 (8.9) days overall. Index SRE hospitalizations were longest for patients discharged to LTC hospitals, mean (SD) 16.6 (13.4) days, followed by patients discharged to SNFs, mean (SD) 9.4 (7.0) days. Table 2 Length of Stay for the SRE Index Hospitalizations by Study Cohorts. Patients Bone Metastases from Solid Tumors Multiple Myeloma Facility admission type after discharge from SRE-related hospitalization n Mean SD Median IQR n Mean SD Median IQR All 7988 8.2 8.9 6 (4,10) 4277 8.0 7.6 6 (4,10) Skilled nursing facility 2629 9.4 7.0 7 (5,12) 1535 8.9 7.0 7 (4,11) Rehabilitation facility 588 8.0 5.4 7 (4,10) 412 8.2 6.8 6 (4,10) Hospice 1076 9.3 13.2 7 (4,12) 306 9.8 9.6 7 (4,13) HHA 1094 8.0 12.6 6 (4,10) 780 7.3 6.9 5 (3,9) LTC hospital 47 16.6 13.4 12 (7,18) 30 21.1 17.1 14.5 (9,28) LTC nursing home 904 8.4 6.7 7 (4,10) 65 11.3 16.7 6 (3,10) Re-hospitalizations within 30 days after SRE discharge 754 7.0 6.3 5 (3,9) 442 6.9 7.5 5 (3,8) Re-hospitalizations after 30 days of SRE discharge 831 5.5 5.3 4 (3,7) 495 5.7 5.6 4 (2,7) HHA, home health agency; IQR, inter-quartile range; LTC, long-term care; SD, standard deviation; SRE, skeletal related event.

For patients with MM, mean (SD) age was 76.6 (7.1) years; 55.7% were women, 82.5% white, and 13.3% black. Mean (SD) length of pre-index SRE-related hospitalization (including the index SRE hospitalization) at baseline was 10.6 (10.4) days and mean (SD) CCI 5.6 (3.3). As shown in Table 2 , the mean (SD) length of index SRE hospitalization was 8.0 (7.6) days overall. Index SRE hospitalizations were longest for patients discharged to LTC hospitals, mean (SD) 21.1 (17.1) days, followed by patients discharged to LTC nursing homes, mean (SD) 11.3 (16.7) days.

Discharge status from the index SRE hospitalization

As shown in Fig. 3 , the largest proportion of patients from the bone metastases cohort were discharged to SNF (32.9%), followed by HHA (13.7%), hospice (13.5%), and LTC (11.3%). The pattern was similar for the MM cohort (SNF, 35.9%; HHA, 18.2%; hospice, 7.2; LTC, 1.5%. Almost 10% of both cohorts were re-hospitalized within 30 days of discharge. Fig. 3 Number and percentage of patients discharged to different facilities after the SRE index hospitalization by study cohort. BM, bone metastasis; MM, multiple myeloma; SRE, skeletal-related event; ST, solid tumor.

Length of stay at each admitting facility after discharge from index SRE hospitalization

Median and inter-quartile range of lengths of stay at each admitting facility after discharge from index SRE hospitalizations are shown in Fig. 4 for both study cohorts. For bone metastases patients, the longest length of stay was for HHA, with a median of 44 days, followed by SNF (median 27 days), and LTC hospital (median 21 days). For MM patients, the longest length of stay was for HHA (median 41 days) followed by LTC hospital (median 23 days), and SNF (median 21 days). Fig. 4 Distribution of length of stay in each facility after discharge from a SRE-related index hospitalization by study cohorts. BM, bone metastasis; MM, multiple myeloma; SRE, skeletal-related event; ST, solid tumor.

Medicare cost at each admitting facility after discharge from index SRE hospitalization

Mean Medicare costs per patient for each admitting facility after discharge from the index SRE hospitalization are presented in Fig. 5 . Except for HHA, the primary cost (>80%) was for the facility itself. For bone metastases patients, the most expensive facility was LTC hospital (mean cost per patient $46,479), followed by rehabilitation facilities (mean cost $25,312), and SNFs (mean cost $23,801). Mean cost per patient for other facilities (not including hospice, mean $6781) ranged from $15,864 to $21,798. For MM patients, the most expensive facility was LTC hospital (mean cost per patient of $49,729), followed by rehabilitation facilities (mean cost $21,242). Mean costs for other facilities (not including hospice, mean $8494) ranged from $15,517 to $19,510. Detailed mean, median, and SD costs per patient per facility overall and by each service category of services are reported in Table 3 . Fig. 5 Medicare cost for each facility after discharge from a SRE-related index hospitalization by study cohorts, mean Medicare allowable per patient per facility, in 2015 $. BM, bone metastasis; LTC, long-term care; MM, multiple myeloma; OP, outpatient; SRE, skeletal-related event; ST, solid tumor. Table 3 Medicare Cost of Each Facility After Discharge from a SRE-related Index Hospitalization by Source of Cost and by Study Cohorts, Mean Medicare Allowable Per Patient Per Facility, in 2015 $. Patients Bone Metastases from Solid Tumors Multiple Myeloma Facility admission type after discharge from a SRE-related hospitalization Facility cost ($) OP visit cost ($) Physician & DME cost ($) Pharmacy cost ($) Total cost ($) Facility cost ($) OP visit cost ($) Physician & DME cost ($) Pharmacy cost ($) Total cost ($) Mean: Medicare allowable amount ($) SNF 19,244 1426 2511 621 23,801 14,590 1,519 2023 729 18,861 Rehabilitation facility 22,562 118 2464 168 25,312 18,663 118 2056 405 21,242 Hospice 6,173 84 416 108 6781 7227 110 431 727 8494 HHA 4779 4245 6675 2402 18,101 3846 4555 5746 4004 18,151 LTC hospital 40,752 222 5128 377 46,479 42,043 531 5471 1683 49,729 LTC nursing home 15,436 2088 2117 2156 21,798 8037 2,118 2283 3079 15,517 Re-hospitalizations within 30 days after SRE discharge 14,419 98 2411 80 17,007 15,846 141 2528 222 18,737 Re-hospitalizations after 30 days of SRE discharge 13,312 160 2289 102 15,864 16,803 108 2275 323 19,510 Median: Medicare allowable amount ($) SNF 13,524 137 1565 0 16,718 10,984 0 1251 0 14,367 Rehabilitation facility 20,114 0 1942 0 22,312 18,152 0 1743 0 20,640 Hospice 3145 0 341 0 3548 3612 0 303 0 3984 HHA 3227 1024 2345 50 9632 3014 834 2176 144 11,488 LTC Hospital 34,159 0 3730 0 39,917 35,801 0 3888 0 41,512 LTC nursing home 1880 0 435 15 3360 2820 354 1076 8 4585 Re-hospitalizations within 30 days after SRE discharge 10,327 0 1683 0 12,243 10,484 0 1643 0 12,734 Re-hospitalizations after 30 days of SRE discharge 9698 0 1607 0 11,583 10,789 0 1616 0 13,305 SD: Medicare allowable amount ($) SNF 18,039 3184 3036 4488 22,277 12,523 4,981 2,845 3084 16,895 Rehabilitation facility 13,488 479 2080 834 14,931 7709 843 1,600 1735 8671 Hospice 9949 775 506 754 10,449 12,977 801 734 10,643 19,921 HHA 5107 11,026 12,731 8864 25,443 3697 11,346 9314 10,736 22,422 LTC Hospital 33,686 702 4362 1504 36,832 25,803 1768 5048 3571 29,230 LTC nursing home 35,768 7969 5933 11,893 51,732 15,925 3568 4645 17,028 34,706 Re-hospitalizations within 30 days after SRE discharge 16,513 428 2656 531 18,417 15,686 704 2684 1244 17,874 Re-hospitalizations after 30 days of SRE discharge 18,613 880 2459 622 20,409 16,787 549 2463 1482 18,036 HHA, home health agency; LTC, long-term care; SNF, skilled nursing facility; SRE, skeletal-related event.

Total Medicare costs by discharge facility are reported in Table 4 , with ranges from $2.18 million (LTC hospital) to $62.57 million (SNF) for bone metastases patients, and from $1.01 million (LTC nursing home) to $28.95 million (SNF) for MM patients. Estimated total Medicare costs for all facilities combined were $152.46 million for 7923 bone metastases patients and $74.69 million for 4054 MM patients. Table 4 Total Medicare Allowable Cost in 2015 $ of Each Facility After Discharge from a SRE Index Hospitalization by Study Cohorts. Total Medicare allowable amounts ($) Facility admission type after discharge from a SRE-related hospitalization n of SRE-related IP claims * Total Length of Stay (days) Facility cost ($) OP visit cost ($) Physician & DME cost ($) Part D pharmacy cost ($) Total cost ($) Patients with Bone Metastases from Solid Tumors Skilled nursing facility 2629 106,522 50,593,606 3,747,784 6,600,707 1,631,537 62,573,635 Rehabilitation facility 588 9093 13,266,579 69,184 1,448,861 98,815 14,883,439 Hospice 1076 27,833 6,642,439 90,165 447,774 115,913 7,296,291 HHA 1094 74,081 5,228,396 4,643,555 7,301,941 2,628,212 19,802,104 LTC Hospital 47 1326 1,915,340 10,451 241,030 17,715 2,184,536 LTC nursing home 904 59,379 13,954,065 † 1,887,954 1,914,208 1,949,162 19,705,388 Re-hospitalizations within 30 days after SRE discharge 754 5,221 10,871,590 73,845 1,817,735 60,377 12,823,548 Re-hospitalizations after 30 days of SRE discharge 831 4,643 11,062,507 132,977 1,902,287 84,888 13,182,659 Patients with Multiple Myeloma Skilled nursing facility 1,535 47,920 22,396,049 2,331,456 3,104,533 1,119,182 28,951,219 Rehabilitation facility 410 5,088 7,651,718 48,376 843,040 166,186 8,709,321 Hospice 306 9,600 2,211,311 33,581 131,758 222,486 2,599,136 HHA 779 45,469 2,995,893 3,548,484 4,476,284 3,118,916 14,139,578 LTC Hospital 30 856 1,261,299 15,941 164,138 50,486 1,491,864 LTC nursing home 65 2,223 522,405 137,653 148,406 200,141 1,008,606 Re-hospitalizations within 30 days after SRE discharge 437 3,175 6,924,657* 61,639 1,104,629 97,013 8,187,938 Re-hospitalizations after 30 days of SRE discharge 492 3,307 8,267,319 53,252 1,119,476 158,818 9,598,864 HHA, home health agency; LTC, long-term care; SRE, skeletal-related event. * Excluding those with total cost of zero. † Total LTC nursing facility cost was estimated through multiplying days of LTC stay by 2015 US average daily nursing home cost. Patient’s daily LTC costs were $235 per day, average of daily semi-private and daily private room. Source: http://skloff.com/cost-of-long-term-care-by-state-2015/

Discussion

In this study, we examined post-discharge healthcare resource use and costs for patients with SRE-related hospitalizations in two cohorts: elderly patients with bone metastases from solid tumors and elderly patients with MM, using Medicare 20% sample data for 2011–2015. Percentages of patients discharged to institutional post-acute care (i.e., SNF and rehabilitation facility) were 40.3% (32.9% SNF and 7.4% rehabilitation) and 45.5% (35.9% SNF and 9.6% rehabilitation) for patients with bone metastases and MM, respectively. These percentages are higher than in the general Medicare population, reported to be 26.3% in 2015 [15] . Percentages of discharges to SNF alone in our study were 32.9% for bone metastases patients and 35.9% for MM patients. In the general Medicare population, the percentage discharged to SNFs from hospital was reported to be about 20% in 2016 [21] . Percentages of discharges to rehabilitation facilities alone in our study were 7.4% for bone metastases patients and 9.6% for MM patients. In the general Medicare population, the percentage was about 4% in 2015 [22] . In addition, about 9%-10% of patients were readmitted to hospital within 30 days of discharge from the index SRE hospitalization. The mean length of index SRE hospitalizations was about 8 days for the two cohorts, longer than in the general elderly Medicare population, reported as 5.8 days in 2015 [23] . Post-discharge SNF stays were much longer than in general Medicare population: 40.5 days for bone metastases patients and 31.2 days for MM patients, versus 26.5 days in 2015 for elderly Medicare patients [23] .

We estimated post-discharge costs after an index SRE hospitalization for the two study cohorts. For bone metastases patients and for MM patients, the most expensive post-discharge care took place at SNFs. Total Medicare allowable costs at SNFs after discharge from the index SRE hospitalization for both cohorts combined was about $91.52 million, about 40.3% of total post-discharge care. The extrapolated estimate of the same Medicare expenditures for the same patients in the entire Medicare population would be at least $457.6 ($91.52*5) million in the study period. The mean per patient cost in these patient populations after discharge from a SRE-related hospitalization was higher than for the general Medicare population across all facilities. The mean costs at SNFs were $23,801 for bone metastases patients and $18,861 for MM patients. In the general elderly Medicare population, the mean cost per SNF admission in 2015 was $11,545 [23] . The mean costs for readmission within 30 days were $17,007 for bone metastases patients and $18,737 for MM patients. In the general elderly Medicare population, the mean cost per discharge in 2015 was $11,750 [23] . Mean (median) costs at rehabilitation facilities were $25,312 ($22,312) for bone metastases patients and $21,242 ($20,640) for MM patients, versus the median cost of $11,124-$19,443 in the 2015 in general Medicare population [22] . Mean (median) costs at LTC hospitals were $46,479 ($39,917) for bone metastases patients and $49,729 ($41,512) for MM patients, versus the mean cost of $40,718 in the 2015 general Medicare population [22] .

This study has strengths and limitations. To our knowledge, it is the first study to examine discharge status and post-discharge spending for cancer patients subsequent to SRE hospitalization, an important cost component to consider. Medicare is a large US federal government program; among adults aged 65 years or older, 93.7% were insured under a government plan (primarily Medicare) [24] . Therefore, our study results are representative of the US Medicare population with the same disease conditions. Regarding limitations, first, these results are limited to Medicare beneficiaries and cannot be extrapolated to the US general population. Second, this study used claims-based algorithms to define bone metastases, MM, and SREs. Although these definitions have been widely used in previous studies, the entire medical history cannot be obtained from claims data; therefore, the case estimates may be imperfect.

In conclusion, this study provides estimates of discharge status and post-discharge costs after an index SRE hospitalization for elderly Medicare patients with bone metastases from solid tumors and patients with MM in the study period 2011–2015. Compared with the general Medicare population, in the study samples, percentages of admissions to institutional post-acute care facilities such as SNFs, rehabilitation facilities, or LTC facilities were higher, stays in those facilities were longer, and resulting total Medicare costs were higher. Most elderly patients with SRE-related hospitalizations (>75%) were discharged to subsequent healthcare facilities, associated with substantial costs. Post-discharge management after SRE hospitalization is clinically and economically significant for elderly patients, has significant implications for public health, and highlights the need for primary prevention of these painful and costly events.

CRediT authorship contribution statement

Suying Li: Conceptualization, Methodology, Writing - original draft, Writing - review & editing. Haifeng Guo: Formal analysis, Writing - review & editing. Yi Peng: Formal analysis, Methodology, Writing - review & editing. Tingting Gong: Formal analysis, Writing - review & editing. Alan Fu: Conceptualization, Methodology, Writing - review & editing. Debajyoti Bhowmik: Conceptualization, Writing - review & editing. Rohini K. Hernandez: Conceptualization, Writing - review & editing. Katherine B. Carlson: Conceptualization, Writing - review & editing. Kimberly A. Lowe: Conceptualization, Writing - review & editing. Jitesh Rana: Conceptualization, Writing - review & editing. Shuling Li: Conceptualization, Methodology, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

1 Hernandez R.K. Adhia A. Wade S.W. O'Connor E. Arellano J. Francis K. Alvrtsyan H. Million R.P. Liede A. Prevalence of bone metastases and bone-targeting agent use among solid tumor patients in the United States Clin Epidemiol 7 2015 335 345 26229504 10.2147/CLEP.S85496 PMC4514316

2 V. Chia, K. Cetin, J. Jacobsen, M. Nørgaard, A. Jensen, C. Christiansen, H. Sørensen, The incidence and prognostic significance of bone metastases and skeletal-related events in lung cancer patients: A population-based cohort study in Denmark, J Clin Oncol 28 (suppl) (2010) Abstract e18074.

3 Norgaard M. Jensen A.O. Jacobsen J.B. Cetin K. Fryzek J.P. Sorensen H.T. Skeletal related events, bone metastasis and survival of prostate cancer: a population based cohort study in Denmark (1999 to 2007) J Urol 184 1 2010 162 167 20483155 10.1016/j.juro.2010.03.034

4 M. Yong, A. Jensen, J.B. Jacobsen, M. Norgaard, J.P. Fryzek, H.T. Sorensen, Survival associated with bone metastases and skeletal-related events in breast cancer patients: a population-based cohort study in Denmark (1999-2007), J Clin Oncol 27 (suppl) (2009) Abstract e22210.

5 E. Terpos, N. Kaneallias, L.A. Moulopoulos, D. Christoulas, A. Koureas, T. Bagratuni, V. Koutoulidis, E. Kastritis, M.A. Dimopoulus, Skeletal-Related Events In Patients With Multiple Myeloma In The Era Of Novel Agents: Low Incidence Of Pathological Fractures After Treatment, 2013. [abstract] Available at: https://ashpublications.org/blood/article/122/21/3090/11606/SkeletalRelated-Events-In-Patients-With-Multiple. Accessed March 16, 2020.

6 National Cancer Institute, SEER Cancer Stat Facts: Myeloma. 2014. Available at: https://seer.cancer.gov/statfacts/html/mulmy.html. Accessed March 16, 2020.

7 Barlev A. Song X. Ivanov B. Setty V. Chung K. Payer costs for inpatient treatment of pathologic fracture, surgery to bone, and spinal cord compression among patients with multiple myeloma or bone metastasis secondary to prostate or breast cancer J Manag Care Pharm 16 9 2010 693 702 21067255 10.18553/jmcp.2010.16.9.693 PMC10437882

8 Hagiwara M. Delea T.E. Chung K. Healthcare costs associated with skeletal-related events in breast cancer patients with bone metastases J Med Econ 17 3 2014 223 230 24494707 10.3111/13696998.2014.890937

9 Hagiwara M. Delea T.E. Saville M.W. Chung K. Healthcare utilization and costs associated with skeletal-related events in prostate cancer patients with bone metastases Prostate Cancer Prostatic Dis 16 1 2013 23 27 23146970 10.1038/pcan.2012.42

10 Bhowmik D. Hines D.M. Intorcia M. Wade R.L. Economic burden of skeletal-related events in patients with multiple myeloma: analysis of US commercial claims database J Med Econ 21 6 2018 622 628 29571273 10.1080/13696998.2018.1457531

11 Nash Smyth E. Conti I. Wooldridge J.E. Bowman L. Li L. Nelson D.R. Ball D.E. Frequency of skeletal-related events and associated healthcare resource use and costs in US patients with multiple myeloma J Med Econ 19 5 2016 477 486 26671598 10.3111/13696998.2015.1132225

12 McDougall J.A. Bansal A. Goulart B.H. McCune J.S. Karnopp A. Fedorenko C. Greenlee S. Valderrama A. Sullivan S.D. Ramsey S.D. The Clinical and Economic Impacts of Skeletal-Related Events Among Medicare Enrollees With Prostate Cancer Metastatic to Bone Oncologist 21 3 2016 320 326 26865591 10.1634/theoncologist.2015-0327 PMC4786354

13 Svendsen M.L. Gammelager H. Svaerke C. Yong M. Chia V.M. Christiansen C.F. Fryzek J.P. Hospital visits among women with skeletal-related events secondary to breast cancer and bone metastases: a nationwide population-based cohort study in Denmark Clin Epidemiol 5 2013 97 103 23576882 10.2147/CLEP.S42325 PMC3616605

14 Ponnusamy K.E. Naseer Z. El Dafrawy M.H. Okafor L. Alexander C. Sterling R.S. Khanuja H.S. Skolasky R.L. Post-Discharge Care Duration, Charges, and Outcomes Among Medicare Patients After Primary Total Hip and Knee Arthroplasty J Bone Joint Surg Am 99 11 2017 e55 10.2106/JBJS.16.00166 28590385

15 Werner R.M. Konetzka R.T. Trends in Post-Acute Care Use Among Medicare Beneficiaries: 2000 to 2015 JAMA 319 15 2018 1616 1617 29677292 10.1001/jama.2018.2408 PMC5933328

16 Li S. Peng Y. Weinhandl E.D. Blaes A.H. Cetin K. Chia V.M. Stryker S. Pinzone J.J. Acquavella J.F. Arneson T.J. Estimated number of prevalent cases of metastatic bone disease in the US adult population Clin Epidemiol 4 2012 87 93 22570568 10.2147/CLEP.S28339 PMC3345874

17 Princic N. Gregory C. Willson T. Mahue M. Felici D. Werther W. Lenhart G. Foley K.A. Development and validation of an slgorithm to identify patients with multiple myeloma using sdministrative claims data Front Oncol 6 2016 224 27833899 10.3389/fonc.2016.00224 PMC5081355

18 Aly A. Onukwugha E. Woods C. Mullins C.D. Kwok Y. Qian Y. Arellano J. Balakumaran A. Hussain A. Measurement of skeletal related events in SEER-Medicare: a comparison of claims-based methods BMC. Med. Res. Methodol 15 2015 65 26286392 10.1186/s12874-015-0047-5 PMC4544826

19 Centers for Medicaid & Medicare Services. Patient Discharge Status Codes Matter. Available at: https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/Fast-Facts/Patient-Discharge-Status-Codes-Matter.html. Accessed March 16, 2020.

20 Skloff Financial Group. Cost of long-term care by state. Available at: http://skloff.com/cost-of-long-term-care-by-state-2015/. Accesssed March 16, 2020.

21 Medicare Payment Advisory Committee. Report to the Congress: Medicare Payment Policy 2019: Skilled nursing facility services. Available at: http://medpac.gov/docs/default-source/reports/mar19_medpac_ch8_sec.pdf. Accesssed March 16, 2020.

22 Medicare Payment Advisory Committee. Health Care Spending and the Medicare Program, June 2017. Available at: http://medpac.gov/docs/default-source/data-book/jun17_databookentirereport_sec.pdf. Accesssed March 16, 2020.

23 Centers for Medicare & Medicaid Servicdes. Medicare utilization and payment. Available at: https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/CMSProgramStatistics/2015/2015_Utilization. Accesssed March 16, 2020.

24 E.R. Berchick, E. Hood, J.C. Barnett, Health Insurance Coverage in the United States: 2017. In: Current Population Reports, 2017., Washington, DC, 2018, 2018. Available at: https://www.census.gov/content/dam/Census/library/publications/2018/demo/p60-264.pdf. Accessed on March 16, 2020.

Appendix A International Classification of Diseases, Ninth Edition, Clinical Modification diagnosis codes used to identify solid tumors. Cancer Types Diagnosis Codes Head and neck 140.x-149.x Esophagus 150.x Stomach and small intestine 151.x, 152.x Colon and rectum 153.x, 154.x Liver 155.x Gallbladder 156.x Pancreas 157.x Retroperitoneum and peritoneum 158.x Spleen 159.x Lung 162.x Other respiratory 160.x-161.x, 163.x, 164.x-165.x Bone 170.x Connective and soft tissue 171.x Melanoma 172.x Female breast 174.x Male breast 1750 or 1759 Sarcoma 176.x Gynecologic 179.x-184.x Prostate 185 Other genitourinary 186.x-189.x Central nervous system 190.x-192.x Endocrine 193.x, 194.x Other and ill defined 195.x, 199.x Appendix B International Classification of Diseases, Ninth Edition, Clinical Modification diagnosis codes used to identify multiple myeloma. Description Code Multiple myeloma without mention of having achieved remission 203.00 Multiple myeloma in remission 203.01 Multiple myeloma in relapse 203.02 Appendix C Codes and algorithm to identify skeletal-related events (SREs). Event ICD-9/CPT/HCPCS Codes Number/Type of Claims to Qualify for Case Definition Skeletal Related Events (SRE) (will use algorithm of “base case” definition from Aly et al.) Spinal Cord Compression: ICD-9 diagnosis codes: 336.9 (unspecified disease of the spinal cord) ICD-10 diagnosis codes: G061, G07, G9520, G9529, G959 HCPCS codes: 63050, 63051, 22551, 22552, 63064, 63066, 61343, S2348, 63075–8, S2350, S2351, 63195, 63197, 63199, 63001, 63003, 63005, 63011, 63015, 63016, 63017, 63170, 63012, 63045, 63046, 63047, 63048, 63040, 63042, 63043, 63044, 63020, 63030, 63035, 22224, 22222, 22214, 22212, 22207, 22206, 0274 T, 0275 T, C9729, 0202 T, 22865, 0164 T, 0094 T, 0097 T, 63057, 63056, 63055, 63081, 63082, 63087, 63088, 63101, 63102, 63103, 63090, 63091, 63086, and 63,085 1 Part A or 1 Part B Pathologic Fracture: ICD-9 diagnosis codes for pathologic fracture: 733.1X (pathologic fracture) ICD-10 diagnosis codes for pathologic fracture: not listed in this table. ICD-9 diagnosis codes for other fracture: 8202, 8208, 8210, 8212, 8120, 8122, 8124, 8130, 8132, 8134, 8138, 8230, 8232, 8238, 805, 806, 8200, 800, 807, 8080, 8082, 8084, 8088, 8100, 8240, 8242, 80841, 80842, 80843, and 80849; ICD-10 diagnosis codes for other fracture: not listed in this table ICD-9 diagnosis codes for trauma/non-routine falls/accidents: 819, 828, 851, 852, 853, 854, 860, 861, 862, 863, 864, 865, 866, 867, 868, 869, 9584, E800-E848, E881, E882, E883, E884.0, E884.1, E884.5, E885.0, E885.1, E885.2, E885.3, E885.4, E886.0, E886.9, E888.0, and E888.1 ICD-10 diagnosis codes for Trauma: not listed in this table 1 Part A or 1 Part B indicating “pathologic fracture” and “other fracture” excluding codes suggesting concurrent accidents/fall defined as within 2 weeks ending with the fracture Bone Palliative Radiotherapy: ICD-9 procedure codes: 92.23, 92.24, 92.29 ICD-10 procedure codes for Radiotherapy: not listed in this table HCPCS codes: A9600, A9604, A9605, C9401, J3005, 77401, 77402, 77403, 77404, 77406, 77407, 77408, 77409, 77411, 77412, 77413, 77414, 77416, 77418, 79005, 79101, 79200, 79300, 79400, 79403, 79440, 79445, 79999, and new from 2015 including G6003-G6015 (for radiation in office) and 77385–77386 (for hospital outpatient) and HCPCS codes: G6003-G6015 and CPT codes: 77,385 and 77,386 1 Part A or 1 Part B claim indicating receipt of EBRT (2D and 3D) or radioisotopes Bone Surgery: ICD-9 procedure codes: 7815, 7845, 7855, 7915, 7925, 7935, 7995, 7812, 7842, 7852, 7911, 7921, 7931, 7991, 7813, 7843, 7853, 7912, 7922, 7932, 7992, 7817, 7847, 7857, 7916, 7926, 7936, 7996, 0353, 8102, 8103, 8104, 8105, 8106, 8107, 8108, 7810, 7811, 7816, 7819, 7840, 7841, 7846, 7849, 7850, 7851, 7856, 7859, 7910, 7919, 7920, 7929, 7930, 7939, 7990, and 7999 ICD-10 procedure codes for surgery: not listed in this table CPT codes: 27187, 27235, 27236, 27244, 27245, 27248, 27269, 27495, 27506, 27507, 27509, 27511, 27513, 27514, 23615, 23616, 23630, 24498, 24515, 24516, 24538, 24545, 24546, 24566, 24575, 24579, 24582, 24586, 24587, 24635, 24665, 24666, 24685, 25490, 25491, 25492, 25515, 25525, 25526, 25545, 25606, 25607, 25608, 25609, 27535, 27536, 27745, 27756, 27758, 27759, 27766, 27769, 27784, 27792, 27826, 27827, 22325, 22326, 22327, 22328, 22520, 22521, 22522, 22532, 22533, 22534, 22548, 22550, 22554, 22555, 22556, 22558, 22565, 22585, 22590, 22595, 22600, 22610, 22612, 22614, 22615, 22625, 22630, 22632, 20982, 23490, 23515, 23585, 27215, 27216, 27217, 27218, 27226, 27227, 27228, 27524, 27540, 22523, 22524, 22525, 22526, 22527, 25574, and 25,575 1 Part A or 1 Part B indicating any bone surgical procedure

Acknowledgments

The authors thank Chronic Disease Research Group colleague Nan Booth, MSW, MPH, ESL, for manuscript editing.

Funding

This work was supported by Amgen, Inc., Thousand Oaks, California.