Published 2/1/2019
Terry Stanton

Study Examines Medicare Hospital Costs for Spinal Fusion

A study seeking to identify cost drivers among Medicare beneficiaries (MBs) undergoing one- or two-level lumbar spinal fusion (1-2-LLSF) identified 22 demographic factors, comorbid conditions, and fusion approaches that increased costs during hospital stays by more than 5 percent.

The study, presented at the North American Spine Society Annual Meeting by Kevin J. McGuire, MD, MS, of Dartmouth-Hitchcock Medical Center in Lebanon, N.H., considered all U.S. hospitalizations in which a Medicare fee-for-service beneficiary underwent a 1-2-LLSF during the first nine months of 2014. The researchers identified 43,425 hospital admissions with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure code of 81.06, 81.07, or 81.08, indicating that the patient underwent a lumbar fusion procedure (anterior technique, lateral transverse process technique, or posterior technique) and that the fusion procedure included two or three vertebra (ICD-9-CM code 81.62, without having had a refusion procedure) during hospitalization. Exclusion criteria were:

death during the index 1-2-LLSF hospitalization (n = 53)

death within 90 days of the day of discharge (n = 178)

hospitalization with a low estimated hospital total cost (total cost of less than $4,000) (n = 77)

The final study data set consisted of 43,117 hospital admissions in 1,877 U.S. hospitals.

The investigators explored several demographic variables, comorbid condition variables, and fusion approaches for every beneficiary in the study. Demographic variables of interest included age group (younger than 65 years, 65–69 years, 70–74 years, 75–79 years, and 80 years and older), sex, and race (white, black, Hispanic, or all other races). Comorbid conditions were coded as one if the condition was present and zero if the condition was not present (Table 1). Seven fusion procedure approaches were identified:

  1. posterior only (posterior column/posterior technique) (e.g., posterolateral fusion)
  2. anterior only (anterior column/anterior technique) (e.g., anterior lumbar interbody fusion [ALIF] or lateral lumbar interbody fusion [LLIF])
  3. anterior column/posterior technique only (e.g., transforaminal lumbar interbody fusion [TLIF])
  4. posterior and anterior column/posterior technique (e.g., posterolateral and TLIF)
  5. anterior and posterior (e.g., ALIF or LLIF and posterolateral)
  6. anterior and anterior column/posterior technique (e.g., ALIF or LLIF and TLIF)
  7. anterior, posterior, and anterior column/posterior techniques (e.g., ALIF or LLIF, posterolateral, and TLIF)

The investigators estimated total hospital costs during patients’ 1-2-LLSF hospital admissions; the estimates do not include other healthcare resources provided by physicians or other postacute facilities. Hospital total costs were estimated by multiplying total hospital charges that were found in the Inpatient Standard Analytical Files (Medicare claims) by the appropriate hospital’s overall cost-to-charge ratio, obtained from the fiscal year 2014 Medicare Cost Report.

Univariate statistics were used to describe demographics, comorbid conditions, and fusion approaches. The authors reported all observed event rates as the proportion of 1-2-LLSF hospitalizations with a selected condition out of all study hospitalizations. Mean hospital costs for all patients are presented as mean ± standard deviation (SD).

Most patients undergoing a 1-2-LLSF were white (88.8 percent), female (59.4 percent), and aged 65–74 years (Table 1). Compared to males, female patients undergoing a 1-2-LLSF were significantly more likely to be younger than 65 years (21.2 percent versus 20.0 percent, P = 0.004) and nonwhite (11.8 percent versus 10.2 percent, P < 0.001). Conversely, male patients undergoing a 1-2-LLSF were significantly more likely to be white (89.8 percent versus 88.2 percent, P < 0.001) and aged 75 years and older (26.7 percent versus 25.7 percent, P = 0.019).

On average, MBs undergoing a 1-2-LLSF had three (SD = 2.0) comorbid conditions present during 1-2-LLSF hospitalization. The three most common comorbid conditions were hypertension (69.0 percent), diabetes mellitus (25.2 percent), and degenerative disk disease (20.0 percent). A total of 3,131 MBs (7.3 percent) undergoing a 1-2-LLSF had none of the study’s comorbid conditions. Female MBs were significantly more likely to not have any of the comorbid conditions (7.6 percent versus 6.8 percent,
P < 0.001) compared to male MBs. Male MBs were significantly more likely to have more than five comorbid conditions (14.6 percent versus 9.5 percent, P < 0.001).

Posterior only (posterior column/posterior technique) was used as the fusion approach on more than half (51.4 percent) of MBs undergoing a 1-2-LLSF (Table 2). The next two most common fusion approaches were anterior column/posterior technique only (21.3 percent) and posterior column/posterior technique and anterior column/posterior technique (12.2 percent). Overall, there were no significant differences in the use of fusion approaches between male and female MBs undergoing a
1-2-LLSF (P = 0.127).

Cost factors

Average total cost of hospital resources consumed during 1-2-LLSF hospitalization was $27,182 (SD = $15,569). In addition, costs varied by age category from a low of $25,177 (SD = $14,390) among MBs aged 80 and older to a high of $27,715 (SD = $16,366) among patients younger than 65 years. There was no significant difference in the observed average total hospital cost between female and male MBs ($27,216 versus $27,133, P = 0.589). However, nonwhite MBs had a significantly higher average total hospital cost ($28,607 versus 27,003, P < 0.001) than white MBs. Observed total hospital cost averaged more than $40,000 for three of the fusion combination approaches: ALIF or LLIF and posterolateral fusion ($41,251 [SD = $16,490]); ALIF or LLIF and TLIF ($43,190 [SD = $25,148]); and ALIF or LLIF, posterolateral, and TLIF ($51,427 [SD = $22,889]).

The hospital cost equations estimated that the average cost for a 1-2-LLSF for a white female aged 80 years and older with no comorbid conditions was $21,881. The estimated incremental hospital costs were significantly higher for black (+ $770) and other race (+ $1,517) MBs than white MBs. Estimated incremental hospital costs were significantly higher for all age categories relative to MBs who were 80 years and older, however. MBs aged 65–69 years were estimated to have the highest incremental cost (+ $2,238).

Fourteen of the study’s comorbid conditions were associated with significantly higher hospital costs. Four comorbid conditions were estimated to increase incremental hospital cost by more than $10,000: malnutrition (+ $19,243), malignant tumor present on admission (+ $14,780), bone infection present on admission (+ $13,025), and benign tumor present on
admission (+ $11,875). On the other hand, six comorbid conditions were significantly associated with lower incremental hospital cost; however, only one of those conditions (long-term use of antiplatelets [– $2,173]) was estimated to lower incremental cost by more than $850. Finally, compared to posterior only (posterior column/posterior technique), every other fusion approach was estimated to significantly increase incremental costs by at least $3,106.

The implications

The authors wrote that several important findings emerged from their analysis. First, they noted, more than 2 percent of patients undergoing a 1-2-LLSF were younger than 65 years. “The vast majority of this group of MBs most likely have long-term disabilities, which may explain why this age group had incrementally higher hospital costs ($1,770) than MBs aged 80 years and older,” the authors observed. “In addition, slightly more than half of MBs undergoing a 1-2-LLSF were aged 65–74 years.”

Second, only 7.3 percent of patients (n = 3,131) undergoing a 1-2-LLSF presented without at least one of the 31 comorbid conditions identified. The number of comorbid conditions any MB had at the time of 1-2-LLSF ranged from zero to 13. The eight most common comorbid conditions among MBs undergoing a 1-2-LLSF were hypertension (69.0 percent), diabetes mellitus (25.2 percent), degenerative disk disease (20.0 percent), history of smoking (18.9 percent), chronic obstructive pulmonary disease (18.9 percent), spondylolisthesis (17.2 percent), chronic ischemic heart disease (16.6 percent), and obesity (15.5 percent).

Third, seven different fusion approaches were used to treat MBs undergoing a 1-2-LLSF. More than half of all MBs (51 percent) received the posterior only (posterior column/posterior technique) fusion approach during 1-2-LLSF hospitalization. The next three most common fusion approaches were: (1) anterior column/posterior technique only (21.3 percent), (2) posterior and anterior column/posterior technique (12.2 percent), and (3) anterior only (anterior column/anterior technique) (8.2 percent).

Dr. McGuire said, “As alternative payment models continue to evolve, the variability in resource utilization and predicting this variability within a reasonable degree of certainty will allow providers and hospitals to provide upstream resources to decrease complication rates and facilitate care, as well as develop a sustainable economic model for care at their institution.”

An appropriate follow-up to the study would be to “look at resource utilization across a broader period of time—i.e., 90 days—and evaluate where the greatest magnitude and variability [are] in resource utilization,” he said. “This would also reevaluate the factors that predict utilization and where in the care process the greatest resources were utilized.”

Possible limitations of the investigation include those inherent to administrative databases, as well as lack of external validity and the retrospective nature of the study.

Dr. McGuire’s coauthors are Steven D. Culler, PhD; Kenneth M. Little, MD; Michael Schlosser, MD; David S. Jevsevar, MD, MBA; Kevin G. Shea, MD; and April W. Simon, RN, MSN.

Terry Stanton is the senior science writer for AAOS Now. He can be reached at tstanton@aaos.org.