Thirty- and 90-day Readmissions After Spinal Surgery for Spine Metastases

A National Trend Analysis of 4423 Patients

Aladine A. Elsamadicy, MD; Andrew B. Koo, MD; Wyatt B. David, MS; Cheryl K. Zogg, MSPH, MHS; Adam J. Kundishora, MD; Christopher S. Hong, MD; Gregory A. Kuzmik, MD; Ramana Gorrepati, MD; Pedro O. Coutinho, MD; Luis Kolb, MD; Maxwell Laurans, MD, MBA; Khalid Abbed, MD


Spine. 2021;46(12):828-835. 

In This Article


In this retrospective national readmission study, we found that patients with spinal metastases who underwent surgical intervention had an unplanned 30- and 90-day readmission rate of 24.1% and 13.3%, respectively. Furthermore, we found that patients with 30-day and 90-day readmissions trended to have longer length of stay and greater total hospital costs during initial admission for surgery.

Previous studies have attempted to identify demographic factors associated with readmission following spine surgery for spinal metastases. In a retrospective analysis of 159 patients with metastatic spine disease treated at a tertiary center from 2003 to 2012, Abu-Bonsrah et al[15] found that age >60 years and African-American race were associated with significantly higher rates of readmission, where male sex was not. Similarly, in a retrospective study of 2207 patients with spinal tumors, Karhade et al[16] found that age >67 years was associated with an increased rate of 30-day readmission, whereas minority racial status, sex, and body mass index were not. In a retrospective study of 164 patients with metastatic tumors of the spine treated at a large referral spine center between 2005 and 2011, Schairer et al[13] demonstrated that increased age was a risk factor for higher rates of readmission. In another retrospective study of 397 patients who had undergone surgical resection of spinal tumors, Janjua et al[17] found that younger age and Medicare insurance were significant predictors of 90-day readmission. Interestingly, we found that younger age was the only demographic factor associated with increased rates of 30-day readmission, whereas neither age nor race was associated with increased rates of 90-day readmission. This increased rate of 30-day readmissions in young patients is contrary to the findings of previous spine literature demonstrating high rates of readmission among the elderly due to myriad factors.[18,19] Surgical factors such as increased length of surgical operation and high complexity of the surgical procedure compound with biological factors such as higher rates of medical comorbidities, higher sensitivity to pharmacological agents, and lower physiological reserves.[18,19] Baseline functional limitations and lower levels of social and financial support further exacerbate postoperative outcomes.[18,19] Intriguingly, some studies have found no difference in complication rates,[20] extended LOS,[21] or reoperation rates[21] for octogenarians compared to younger patients following spine surgery, throwing into question the idea that older patients inherently will have worse surgical outcomes. For our study, it is possible that only particularly healthy elderly patients with an extended life expectancy were deemed fit enough to undergo surgical intervention and thus had lower readmission rates. Alternatively, health care providers and family members may be more diligent with postoperative mobilization and recovery efforts for elderly patients leading to fewer returns to the hospital. Regardless, future studies are thus indicated to more fully elucidate the impact of age on postoperative unplanned readmissions.

Along with demographic factors, other studies have examined the impact that comorbidities have on predisposing patients to unplanned readmission. In the Schairer et al[13] study of 164 patients undergoing resection of metastatic spinal tumors, numerous comorbidities were associated with an increased risk of readmission, including pulmonary disorder, diabetes, depression, hypothyroidism, and obesity. Similarly, in the Karhade et al's[16] study of 2207 patients, the authors found an increased risk of 30-day readmission amongst patients with comorbid dyspnea, HTN, and anemia. In contrast, in the Abu-Bonsrah et al's study of 159 patients, the number of comorbidities was not found to be associated with increased risks of unplanned hospital readmission.[15] In the Lau et al's[14] study of 118 patients, the authors found that vascular disease was the only comorbidity associated with significantly increased readmission rates at the 90-day mark. Analogous to the Abu-Bonsrah et al's study, the Lau et al's study also found no association between number of comorbidities and increased rates of readmission.[14,15] Overall, our study found that HTN, renal failure, and rheumatoid arthritis were independently associated with increased rates of unplanned readmission, whereas coagulopathy was interestingly associated with a decreased risk of readmission. The association between chronic conditions and increased readmission rates following spine surgery has been well documented in literature.[22] Indeed, patients with increased comorbidity burden have been shown to have higher levels of complications and morbidity following surgery, predisposing them to unplanned hospital readmission.[23,24] It is possible that the comorbidities themselves may be directly responsible for unplanned readmission, for example with pulmonary dysfunction or deficiency anemia leading to decreased postoperative mobilization.[23,25] However, the comorbidity may instead be a proxy marker for a separate health problem that is the true etiology of the worsened outcome and unplanned readmission.[23] Regardless, controlling these comorbidities preoperatively as well as adequate perioperative management can result in improved patient outcomes and decreased unplanned readmission rates.[25]

Understanding the early prevalence of common postoperative complications following surgical resection of spinal metastases may allow for avenues to be identified in efforts to decreased unplanned readmission. In fact, in the Schairer et al's[13] study of 164 patients, the authors noted that 19.2% of readmissions were due to surgical complications. In a prospective study of 282 patients undergoing resection of spinal metastases, Jansson et al[10] found a 20% complication rate, with the most common complications being cardiac, pulmonary, infection, and hematoma. In the Abu-Bonsrah et al's[15] study of 159 patients, 47.8% of patients developed at least one perioperative complication, with the most common being respiratory complication, excessive pain, and cardiac complication. In a retrospective study of 135 patients undergoing resection of metastatic spinal cord tumors, Elsamadicy et al[9] found a 20% rate of weakness, 16.7% of new sensory deficits, and 13.3% of hypotension postoperatively. In a prospective study of 152 patients undergoing surgery for spinal tumor metastasis at a tertiary care center, Atkinson et al[26] found the rate of surgical site infection to be 11.2%, which was associated with a higher mortality rate. Similarly, in a retrospective study of 141 patients undergoing metastatic spinal tumor resection, Roser et al[27] found a 19.8% rate of major complications, with the most common being wound infection or dehiscence. Analogous to the aforementioned studies, our study identified high rates of postoperative infection and acute post-hemorrhagic anemia in patients undergoing surgical intervention for spinal tumor.

Few studies have found associations between complications rates and unplanned 30- and 90-day readmission. In the Janjua et al's[17] study of 397 patients, the authors found that pulmonary complications, excessive bleeding, and infection were common complications leading to 30-day unplanned readmission. Furthermore, the authors found that CNS, pulmonary, and infectious complications were among the five most common causes for 90-day readmission.[17] Abu-Bonsrah et al[15] found that 26.7% of the unplanned readmissions were due to surgical complications, with the most common being wound infections and cerebrospinal fluid leak. Similarly, in the Elsamadicy et al's[9] study of 135 patients, the authors showed that uncontrolled pain (16.7%), sensory motor deficits (13.3%), and fever (10.0%) were the most common reasons for 30-day readmission. Overall, our study found 30- and 90-day readmission to most commonly be caused by sepsis, postoperative infection, and genitourinary complications. Further investigations are warranted to identify quality initiatives that may be applied to reduce adverse complications to better patient quality of care and reduce unplanned readmissions.

In an era where the United States spends nearly twice as much as other developed countries on health care, reducing unplanned readmissions is a current target to improve care and decrease cost.[28,29] Readmissions in general cost the US health care system $15 billion annually.[30] Lau et al[14] found that patients undergoing readmission following epidural metastasis resection accrued an additional cost of $20,078. This is particularly important because a large number of cancer patients experience bone metastases; 75% of patients with the most prevalent forms of cancer—breast and prostate—experiences bone metastases.[31] Spine metastases are common, with rates found to be as high as 50% in all cancer patients.[32] Hospital readmissions in spinal metastases patients account for the second highest cost of treatment behind index admission, with the entire treatment costing $55,349 on average.[31] Not only do unplanned readmissions increase health care costs, but they also are associated with decreased quality of care.[33,34] Hospitals with high rates of readmission also have high rates of morbidity and mortality.[3,6]

There is a paucity of studies identifying other characteristics associated with increased hospital readmissions. In a retrospective study of 118 adult patients who underwent resection of metastatic tumors of the spine between 2008 and 2013, Lau et al[14] showed that factors such as diagnosis of lung metastasis and index hospitalization length of stay were independently associated with increased rates of 90-day readmission, whereas previous radiation therapy and presence of extraspinal metastases were not. Our study demonstrated that surgery type plays a role in postoperative readmissions, with spinal decompression and fusion having higher rates of 30-day readmission compared to decompression only. Some studies have theorized that the addition of fusion should lead to shortened recovery times, decreased rehabilitation needs, and improved patient outcomes.[35] However, the addition of fusion to a decompression surgery adds instrumentation, operation time, and case complexity to the procedure. Indeed, some previous studies have demonstrated no significant difference in readmission rates at 90 days between decompression and fusion or decompression alone.[36,37] This equivalency has also been shown in randomized clinical trials showing no difference in clinically meaningful outcomes at 2 and 5 years.[38] However, a separate randomized clinical trial demonstrated clinically meaningful improvement in physical health-related quality of life for laminectomy and fusion compared to laminectomy alone.[39] Further studies identifying the impact of surgery on readmission rates are indicated to better optimize patients perioperatively and improve patient care while decreasing health care costs.

This study has several inherent limitations, which has potential implications for its interpretation. First, the analysis is retrospective, with data only available by ICD-9-CM codes which may contain coding and reporting biases. Moreover, although we attempted to control for the majority of comorbidities, there may be other patient factors that may be protective for readmission that are not identified and may have implications on our findings and results. Second, there is a possibility of misclassified or incomplete data. Furthermore, given the nature of large, national database studies, minor differences become statistically significant; therefore, we reported trends instead of statistical differences which may have implications on the interpretation of our results. Thirdly, identifying the sole driver of readmission is not well characterized in the NRD database; therefore, utilizing prevalent complications codes associated with the indexed hospital readmission has interpretation bias and may not be the true reason for the hospital readmission. Lastly, identifying the primary tumor and data describing adjuvant therapy is limited and may have implications on our findings. Regardless, this study has shed light on risk factors that increase readmission rates in patients undergoing surgical intervention for spinal metastases.