Background and rationale
As the number of cases of bone-seeking cancers increases in the United States, the amount of orthopedic oncology services and their associated costs could be a strain on the Medicare system. This study was completed to evaluate the burden of femoral metastases on the Medicare system and to analyze the associated charges.
This study had a number of limitations. First, this is a retrospective observational study that relied on the Medicare inpatient database. All such databases are subject to errors and inaccuracies in coding. For instance, while a prophylactic intramedullary nailing should be coded with 27495, it may have been miscoded with CPT 27506: “Open treatment of femoral shaft fracture, with or without external fixation, with insertion of intramedullary implant, with or without curettage.” Second, the database was not designed to provide more specific patient information such as age, comorbidities, preoperative diagnoses, or postoperative diagnoses, so this limited our ability to gather more detailed information on the population with femoral metastases.
Third, no ICD-9 code that is specific to metastatic disease of the femur exists, so the best surrogate marker was ICD-9 code 198.5—secondary malignant neoplasm of bone or bone marrow. Due to this, we had no way of knowing the percentage of patients with metastatic disease of the femur that did not undergo prophylactic surgical management. However, it is reasonable to assume that the percentage is increasing in the over 65 age groups since the utilization of prophylactic fixation techniques has not increased, but the number of cases of metastatic disease to the skeleton has increased. Fourth, ICD-9 code 78.55 encompasses prophylactic fixation techniques for metastatic disease to the femur, so we were unable to study trends in specific procedures. While more specific CPT codes do exist, these are not used for coding the inpatient Medicare population in the PearlDiver database. Additionally, the data that PearlDiver has available is a couple of years behind the current date because of all the data they compile. In order to get more current information, we used the RBRVS DataManager Online from the American Medical Association to gather data on prophylactic femoral fixation in the years 2013–2014. Due to inaccuracies with coding, CPT 27495 and 27187 may not match the results of ICD-9 78.55 perfectly. Also, we could only extract total volume for these procedures—we did not have access to information regarding hospital charges, Medicare reimbursement, or length of hospital stay.
Moreover, during the statistical analysis, an emphasis on the significance of the trends was placed. This, combined with multiple sub-group analyses, increased our chances of a false positive result.
Finally, since the data we gathered from the Medicare database lags by a couple of years, there is time for trends to change between what the database reflected and what is currently happening in clinical practice.
Discussion: (1) In the Medicare population, has the number of skeletal metastases increased?
The data we collected confirms that there is an increased burden of skeletal metastases in the Medicare population since all groups showed significant upward trends between 2005 and 2012. While the increase in cases does burden the Medicare system, the prevalence of skeletal metastases among Medicare patients has not changed. This reflects the increase in cases of bone seeking cancers, which is what data from the American Cancer Society suggests (American Cancer Society 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012). Also, the increased number of cases can be explained by the increase in the Medicare population (Centers for Medicare and Medicaid Services 2013).
Discussion: (2) In the Medicare population, has the use of prophylactic fixation techniques increased?
Utilization of prophylactic femoral fixation in adults over age 65
With the increased cases of skeletal metastases, the potential need for surgical management of metastatic disease to the skeleton should be increasing. After focusing our attention on the most common site of long bone metastases, we found that prophylactic fixation of the femur, as reported by ICD-9 inpatient procedural code 78.55, CPT 27495, and CPT 27187, has not shown a significant increase in the total Medicare population between 2005 and 2014. Also, the largest of the Medicare subpopulations, adults over the age of 65, did not show an increasing trend between 2005 and 2012. In fact, its rate of use among older adults with skeletal metastases significantly decreased.
Since prophylactic fixation procedures have decreased despite the consistent prevalence of metastatic disease to the skeleton, then there are either fewer metastases to the femur, there are other CPT codes being used, or there are other treatment modalities that are increasingly being utilized to prevent impending fractures. Despite our inability to specifically study the rate of femoral metastases because no ICD-9 code for it exists, we believe that the notion that femoral metastases are decreasing can be dismissed. There is simply no evidence that cancer pathology or therapy have changed in a way that would alter the location preference of a bone-seeking cancer.
Another possible explanation is that prophylactic fixation techniques have not increased because other procedures are being used to stabilize impending fractures. The use of CPT code 27125, “hemiarthroplasty, hip, partial (e.g., femoral stem prosthesis, bipolar arthroplasty)” could explain why prophylactic fixation techniques have not become more common. Studying the rate of CPT 27125 could be an avenue of future research; however, it may be difficult to know whether the procedures were done prophylactically since it is not specified in the code and no prophylactic hemiarthroplasty code exists.
The third explanation, other treatment modalities are stabilizing femoral metastases, is the most plausible. In particular, radiotherapy and osteoclast inhibiting medications have been used to treat metastatic disease to the femur (Bickels et al. 2009). Radiation has been effective for pain management in those with metastatic femoral lesions and was shown to circumvent the need for surgical intervention in 81 % of impending fracture cases in one study (Harada et al. 2010). Other benefits of using radiation therapy over surgery include decreased pain, risk of DVT, fat embolism, anesthesia risk, and hospital stay (Hattori et al. 2007; Kelly et al. 2012; Swanson et al. 2000). Additionally, using radiotherapy instead of surgery may be more cost effective than prophylactic fixation and further studies would be required to evaluate this.
Another non-surgical treatment is the use of bisphosphonates or RANKL inhibitors, which prevent osteoclasts from resorbing bone. They have both been shown to decrease the number of bony metastases, decrease the prevalence of pathologic fractures, and prevent the need for surgical fixation (Bickels et al. 2009; Gartrell and Saad 2014; Saad et al. 2002). While both bisphosphonates and RANKL inhibitors are effective, a recent meta analysis has shown that RANKL inhibitors are better at preventing pathologic fractures than bisphosphonates (Lipton et al. 2012). The increased use of these medications may be circumventing the need for prophylactic fixation. Again, comparing the costs of this therapy with the costs of prophylactic fixation could be a future area of research to determine the most effective use of resources.
Despite a possible increase in the utilization of other treatment modalities, the benefits of early prophylactic fixation have recently been re-emphasized in the literature. One study did so by comparing prophylactic intramedullary nailing outcomes with the outcomes of therapeutic intramedullary nailing after a pathologic fracture. Significant benefits supporting the use of prophylactic nailing included shorter hospital stay, earlier weight bearing, and increased survival (Arvinius et al. 2014). Also, it is important to note that the majority of pathologic fractures never heal, especially if they have previously received radiation therapy (Haidukewych 2012; Miller et al. 2011). This could adversely affect ambulation and put the hardware at a higher risk of failure.
Additionally, prophylactic intramedullary nails have been shown to have a low failure rate, 11 %, and a low complication rate, 12.5 %, which support their continued use (Alvi and Damron 2013). This has led to the conclusion that it is appropriate to protect the entire length of the bone in case of disease progression. However, one study suggests that it is unnecessary to routinely protect the femoral neck since none of their 145 study participants developed metastases in that region (Moon et al. 2014).
Utilization of prophylactic femoral fixation in the under age 65 population
The two groups that did show significant upward trends in the use of prophylactic fixation techniques were total adults under the age of 65 and women under the age of 65. Since men under the age of 65 did not show a significant trend, it is likely that the women in the total population under 65 were the driving force for its significance.
Despite the increase in the number of prophylactic fixation procedures, the rate of their use among those with skeletal metastases did not change in the less than 65 group. It stayed consistent with the prevalence of skeletal metastases. This contrasts with the older population, which saw a decreased trend in the context of an unchanging prevalence.
There has been some evidence that younger women, particularly those less than age 35, have worse breast cancer prognoses (Fredholm et al. 2009; Liu et al. 2015). Since they have more aggressive tumors, perhaps they did not respond well enough to radiotherapy or osteoclast inhibitors and required surgical fixation more often. Also, since they were younger, surgery may have been used more readily since their potential for survival may have been greater or overestimated.
Discussion: (3) How has the financial burden of prophylactic fixation changed over the study period?
Increasing total hospital charges in the under 65 age group
When looking at the estimated total charges, both the overall population’s total hospital charges and the less than 65 age groups had significant upward trends. Because the over 65 age groups did not have significant trends, it is likely that the under 65 age groups were the drivers of the total group’s trend.
A possible explanation for why the younger groups had higher hospital charges could be that those who qualify for Medicare under the age of 65 have significant comorbidities that could be increasing their cost of care.
Significant increase in average hospital charges, but not total hospital charges
All groups except for men under age 65 showed a significant upward trend in the average hospital charges for ICD-9 78.55. It seems contradictory that the total hospital charge for the over 65 age group was not statistically significant, but the average hospital charge was statistically significant. An explanation for this is that the number of procedures has not increased, but the cost per procedure has. This would reflect a higher average hospital charge that may not be reflected in the total hospital charge.
This requires an explanation of why the average charge is increasing after adjusting for inflation. Since there are other treatment modalities in place to treat metastatic disease of the femur, surgery may be seen as more of a final effort to prevent a pathologic fracture. Studies on some surgical procedures, such as pancreaticoduodenectomy, have shown that a decreased volume of procedures can lead to an increase in its cost (Sutton et al. 2014). The decreased use of ICD-9 78.55 could therefore be increasing its value.
While this explanation holds true for the above 65 age groups since they did not have an increase in prophylactic fixation, the less than 65 age group and the women less than 65 age group did have a significant upward trend in prophylactic fixation, so the explanation of fewer procedures yields a higher cost does not apply here. As mentioned previously, the less than age 65 population may have significant comorbidities, which allowed them to enroll in Medicare. The management of those conditions could also be contributing to their increased average hospital charge. For instance, ESRD is one condition that qualifies the under 65 age group for Medicare. Hemodialysis is known to be an expensive procedure and if the patient required dialysis during their hospitalization for prophylactic femoral fixation then this would increase the hospital charge during their stay.
Medicare reimbursement rates
When looking at the total Medicare inpatient reimbursements, only the less than 65 combined age group and the women less than 65 age group showed a significant increasing trend. This reflects the increasing total hospital charges for these groups. Although the men less than 65 age group and the total Medicare inpatient group showed significantly increasing total hospital charges, this was not reflected in their total Medicare reimbursement. The reasons for how Medicare decides their reimbursement rate is beyond the scope of this paper, but our data suggests that it is not solely a proportion of the amount the hospitals decide to charge.
Lastly, the average Medicare reimbursement for ICD-9 78.55 showed a significant upward trend in the women over 65 age group, the combined over 65 age group, and the total Medicare inpatient population. However, all of the significant values were on the lower ends of what is considered a strong correlation and the combined over 65 age group and the total group likely owe their significance to the women over 65 age group. Despite showing a significant increased trend in average hospital charge, the women less than 65 age group, the combined less than 65 age group, and the men over 65 age group did not have a significant positive trend in their average Medicare reimbursement. Again, this suggests that Medicare reimbursement is not simply a proportion of what the hospitals charge.
Length of hospital stay
Even though total and average hospital charges have increased in the total Medicare population, both the over and under 65 year old combined groups, the women over 65 age group, and the total population group showed a significant decreased trend in the average hospital stay. However, since both the men less than 65 and the women less than 65 groups did not show a significant decreased trend, it is possible that the significant trend in the combined less than 65 group is an artifact. The women less than 65 group essentially shows no change in their average hospital stay while the men less than 65 start high at 9 days then drop down to only 6 and 5 days in the last 2 years. This decrease in the last 2 years happened too late in the study period to affect the overall trend in this group, but when averaged with the women under 65 in the combined less than 65 age group, it was able to form a trend.
With regard to the over 65 age group, the decreased trend in average hospital stay was likely driven by the women over 65 group. Since the population in this group was so large, it may have also driven the decreased trend in the total group as well.