Next Article in Journal
Strategies for Maximising Lung Utilisation in Donors After Brain and Cardiac Death: A Narrative Review
Previous Article in Journal
Comparative Study of Methods for Caries Risk Evaluation: CAMBRA, the Cariogram, and Caries Risk Semaphore
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Composite RAI, Malnutrition, and Anemia Model Superiorly Predicts 30-Day Morbidity and Mortality After Surgery for Adult Spinal Deformity

1
Department of Neurosurgery, Zucker School of Medicine at Hofstra, Long Island Jewish Medical Center and North Shore University Hospital, Northwell Health, Manhasset, NY 11030, USA
2
Department of Neurosurgery, Yale University School of Medicine, New Haven, CT 06510, USA
3
Department of Neurosurgery, University of Illinois at Chicago, Chicago, IL 60612, USA
*
Author to whom correspondence should be addressed.
J. Clin. Med. 2025, 14(15), 5379; https://doi.org/10.3390/jcm14155379
Submission received: 18 June 2025 / Revised: 17 July 2025 / Accepted: 28 July 2025 / Published: 30 July 2025
(This article belongs to the Section Orthopedics)

Abstract

Background/Objective: This study examines the composite influence of frailty, malnutrition, and anemia on postoperative outcomes for patients with adult spinal deformity (ASD). Methods: In this retrospective cohort study using the 2011–2022 NSQIP database, we utilized CPT and ICD codes to identify ASD patients who underwent PSF. Subjects were stratified based on frailty status. Frail patients were then classified according to malnutrition and anemia status. Frailty was determined using the revised risk analysis index (RAI-rev). Our primary outcomes were extended length of stay (LOS), non-routine discharge (NRD), 30-day adverse events (AE), and 30-day mortality. For each outcome, we fitted four nested multivariable logistic regression models (RAI-rev + anemia + malnutrition, RAI-rev + anemia, RAI-rev + malnutrition, and RAI-rev alone) and compared the incremental discrimination of each model using receiver operating characteristic (ROC) analysis. Results: Of 3639 patients, 460 were frail alone, 266 were frail + anemic, 37 were frail + malnourished, 121 were frail + anemic + malnourished, and 2755 were not frail. RAI-rev (aOR: 1.84, 95% CI: 1.45–2.35), anemia (aOR: 1.84, 95% CI: 1.45–2.35), and malnourishment (aOR: 2.34, 95% CI: 1.69–3.24) were independent predictors of extended LOS. RAI-rev (aOR: 1.07, 95% CI: 1.04–1.11) and anemia (aOR: 2.09, 95% CI: 1.66–2.61) were associated with an increased risk of 30-day AEs. RAI-rev and malnutrition were independent predictors of NRD (RAI-rev: aOR: 1.11, 95% CI: 1.06–1.16; Malnutrition: aOR: 1.57, 95% CI: 1.08–2.29) and 30-day mortality (RAI-rev: aOR: 1.10, 95% CI: 1.04–1.17; Malnutrition: aOR: 3.79, 95% CI: 1.24–11.60). Based on ROC analysis, RAI-rev + anemic + malnourished was a superior predictor of LOS and 30-day AEs (both p < 0.001). Compared to RAI-rev, RAI-rev + anemic superiorly predicted LOS and 30-day AEs, and RAI-rev + malnutrition superiorly predicted LOS (all p < 0.001). Conclusions: Our results reveal RAI-rev combined with malnutrition and anemia superiorly predicts 30-day AEs and LOS in postoperative ASD patients. Future studies should investigate the feasibility and efficacy of these models for perioperative risk stratification and optimized recovery planning to improve outcomes for ASD patients.

1. Introduction

The past decade has seen an exponential rise in US healthcare costs, with total spending reaching USD 4.9 trillion [1]. Efforts to mitigate expenditures without sacrificing care quality have sparked the development of value-based care protocols [2,3]. Various outcome measures, including hospital length of stay (LOS), discharge disposition, and unplanned readmission, have emerged as proxies for evaluating these protocols [4,5]. In spine surgery patients, extended LOS and unplanned readmission have been linked with increased costs, adverse events (AEs), and mortality [6,7]. One subgroup within spine surgery that would benefit from targeted initiatives to optimize patient outcomes and cost efficiency is adult spinal deformity (ASD) patients. In recent years, there has been an increasing prevalence of ASD [8]—affecting 30–70% of adults older than 60 [9]—and a resulting increase in spinal interventions [8]. These corrective surgical procedures have been associated with higher complication rates, thus increasing the burden on the healthcare system [10,11]. Attempts to optimize perioperative care and decrease costs have sparked an examination of potential risk factors that may influence ASD outcomes.
As the population’s median age increases, one risk factor that has been noted to influence outcomes is frailty. Frailty—decreased adaptability to stressors and lessened reserve of physiologic systems [12]—has been associated with increased mortality, nonroutine discharge (NRD), LOS, readmission, and reoperation [6,13,14,15]. There are various validated indexes, including the Charlson Comorbidity Index [16], the modified frailty index [17], and the recently developed risk analysis index (RAI) [18], which incorporate several comorbidities, such as congestive heart failure (CHF), diabetes, and renal disease, to quantify frailty. In addition to these chronic conditions, frailty often co-occurs with other risk factors, such as anemia [19] and malnutrition [20,21], for adverse surgical outcomes. Despite the increased prevalence of anemia and malnutrition in the setting of frailty, there is a paucity of studies evaluating the combined impact of frailty, anemia, and malnutrition on postoperative ASD outcomes.
This study aims to examine the interplay of frailty, anemia, and malnutrition and their influence on morbidity and mortality in ASD patients undergoing posterior spinal fusion (PSF).

2. Materials and Methods

2.1. Data Source

The American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) is a prospectively collected, peer-controlled database developed for the examination of 30-day risk-adjusted surgical outcomes. Trained personnel abstracted data based on patient records. All data were deidentified, so institutional review board evaluation of this study was not required.

2.2. Patient Cohort

We searched the NSQIP Participant Use Data Files from 2011–2022 for all patients ≥18 years of age who underwent corrective surgery for adult spinal deformity using Current Procedural Terminology (CPT) codes for long fusion-hardware constructs (≥7-level). These CPT codes were 22800, 22802, 22804, 22808, 22810, 22812, 22818, 22819, 22843, 22844, 22846, or 22847. In addition, patients who underwent <7-level fusion (CPT codes 22842 or 22845) were also included if they had a concurrent International Classification of Diseases (ICD) code for spinal deformity, Supplementary Table S1. This approach has been used in the literature to examine this population. [22] Patients who underwent a <7-level spinal fusion without concomitant ICD codes for spinal deformity were excluded. Only procedures performed by neurosurgeons or orthopedic surgeons were included, Figure 1. Study size included all patients meeting inclusion criteria within the timeframe.

2.3. Exposure Definitions

The RAI is a frailty assessment tool that integrates multiple domains, including comorbidities, functional status, nutritional status, and cognitive function. It was recalibrated and validated for surgical populations using large national databases, resulting in the RAI-rev, which improves mortality prediction [23]. As the NSQIP database does not provide all the variables necessary for a complete RAI-rev calculation, we derived the RAI-rev using the available dataset variables as performed in the literature [24] and detailed in Supplementary Table S2. Specifically, we substituted weight loss as a proxy for poor appetite and omitted cognitive decline. The adapted cumulative RAI-rev score ranged from 0 to 78. We utilized RAI-rev to determine frailty in our cohort, classifying participants with scores exceeding 20 as frail, following the criteria set by Conlon et al. [24]. Furthermore, we stratified frail patients based on the presence of anemia and malnutrition. Anemia was defined by preoperative hematocrit levels under 41 for males and under 36 for females. Malnutrition was defined as preoperative serum albumin <3.5 g/dL, a well-established threshold for hypoalbuminemia and nutritional risk [25,26].

2.4. Outcome Definitions

Demographic variables in our analysis included age, sex, race/ethnicity, and body mass index (BMI). Comorbidities included American Society of Anesthesiologists (ASA) grade, diabetes mellitus, hypertension, CHF, chronic obstructive pulmonary disease (COPD), dependent functional status, disseminated cancer, electrolyte abnormalities (defined as preoperative sodium <135 mEq/L or >145 mEq/L), smoking, chronic steroid use, and bleeding disorders. Postoperative AEs occurred within the first 30 days after surgery as provided by NSQIP. These events were categorized into surgical and medical AEs. Surgical AEs included superficial surgical site infection (SSI), organ space SSI, deep SSI, and wound dehiscence. We examined Medical AEs including pneumonia (PNA), need for mechanical ventilation, unplanned reintubation, pulmonary embolism (PE), deep vein thrombosis (DVT), myocardial infarction (MI), C. diff colitis, urinary tract infection (UTI), renal insufficiency, acute renal failure (ARF), systemic sepsis, septic shock, and postoperative red blood cell (RBC) transfusion. AEs were also classified by severity: minor (MAE) and severe (SAE). MAEs included superficial SSI, PNA, renal insufficiency, and UTI, while SAEs included all other AEs. In addition, we examined healthcare utilization outcomes—total operation time (hours), rates of readmission, and incidence of reoperation. Primary outcomes included prolonged LOS (8 days or LOS greater than the 75th percentile for the entire cohort), NRD (discharge to a location other than home or permanent residence), 30-day AEs, and 30-day mortality. We dichotomized LOS using this threshold for ease of interpretability and to avoid issues with skewness and overdispersion that would limit other analytic approaches. Mortality was established using discharge disposition, end-of-life care markers, and death date. Bias was minimized by excluding missing data and with a large-sized sample representative of the larger population.

2.5. Statistical Analysis

The study population was stratified based on frailty status, with frail patients being further categorized according to malnutrition and anemia status. Continuous variables were summarized by mean and SD and categorical variables using frequencies and percentages. Group comparisons utilized ANOVA for normally distributed continuous data, Kruskal–Wallis for non-normal data, and chi-squared or Fisher’s exact tests for categorical data. To evaluate predictors of the study’s primary outcomes, multivariable logistic regression models were created, incorporating RAI-rev, anemia, and malnutrition into each model; we developed four nested logistic-regression models that differed only in the anemia-nutrition variables they contained: (1) RAI-rev + anemia + malnutrition (full model), (2) RAI-rev + anemia, (3) RAI-rev + malnutrition, and (4) RAI-rev alone. For each outcome, the nested models were adjusted for the same covariates, which were determined by considering clinically relevant preoperative and operative variables. Next, the Akaike Information Criterion (AIC) were employed to refine each model in a stepwise fashion, utilizing backward elimination for identifying and retaining the most significant predictors. Adjusted odds ratios (ORs) with 95% confidence intervals were subsequently calculated.
For each outcome, receiver operating characteristic (ROC) curves were created to assess each of the four models’ predictive performance. DeLong tests were then used to compare the area under the curve (AUC), and 95% CIs to quantify the incremental discrimination. Adjusted regression coefficients and corresponding AUCs can be found in the tables and figures, respectively. All tests were two-sided, and the significance was set to p ≤ 0.05. Observations with missing data were excluded (Supplementary Table S3). RStudio v4.4.2 (R Foundation for Statistical Computing, Boston, MA, USA) was utilized for statistical analyses.

3. Results

3.1. Patient Demographics and Comorbidities

Of the 3639 patients identified, 460 (12.7%) were frail alone (F), 266 (7.3%) were frail + anemic (FA), 37 (1.0%) were frail + malnourished (FM), 121 (3.3%) were frail + anemic + malnourished (FAM), and 2755 (75.7%) were not frail (NF). Age (p < 0.001), sex (p < 0.001), and racial/ethnic (p = 0.005) makeup varied significantly, Table 1. NF patients had a lower BMI than patients in the other cohorts (p < 0.001), Table 1. The FAM cohort had the highest proportion of patients with an ASA grade of ≥4 (p < 0.001), Table 1.

3.2. 30-Day Complications and Hospital Outcomes

Frail patients with anemia, malnutrition, or both consistently had worse outcomes than patients in the F cohort. FAM patients had the highest frequency of any AE, reintubation, ventilator requirement, and septic shock (all p < 0.001), Table 2. The FA cohort had the highest incidence of PE (p < 0.001), renal insufficiency (p = 0.004), UTI (p = 0.001), DVT (p = 0.032), C. diff colitis (p = 0.039), and postoperative RBC transfusion (p < 0.001), Table 2. NRD (p < 0.001) was more frequent in FM patients whereas the longest mean LOS (p < 0.001) and highest frequencies of readmission (p = 0.010) and mortality (p < 0.001), Table 2, occurred in FAM patients. Conversely, the NF cohort had the longest total operation time (p = 0.017), Table 2. The predominance of worse outcomes among frail patients with malnutrition, anemia, or both could indicate an increased risk conferred by the combination of these factors compared to frailty alone.

3.3. Multivariable Logistic Regression and ROC Analysis Comparing Predictive Models

3.3.1. Extended LOS

Based on multivariable analysis, RAI-rev-defined frailty (aOR: 1.03, 95% CI: 1.01–1.04), anemia (aOR: 1.84, 95% CI: 1.45–2.35), and malnourishment (aOR: 2.34, 95% CI: 1.69–3.24), Table 3, independently predicted extended LOS. Based on ROC analysis, RAI-rev + anemic + malnourished had the highest AUC of 0.708, indicating moderate discrimination. In other words, given a randomly selected pair of patients, the model has a 71% chance of assigning a higher risk score to the patient with extended LOS. This was followed by RAI-rev + anemic with an AUC of 0.700, RAI-rev + malnourished with an AUC of 0.690, and RAI-rev with an AUC of 0.664, Figure 2A. Compared to RAI-rev, RAI-rev + anemic, RAI-rev + malnourished, and RAI-rev + anemic + malnourished (all p < 0.001) superiorly predicted extended LOS, Figure 2A.

3.3.2. 30-Day Adverse Events

Based on multivariable analysis, risk for 30-day AEs was associated with RAI-rev-defined frailty (aOR: 1.07, 95% CI: 1.04–1.11) and anemia (aOR: 2.09, 95% CI: 1.66–2.61), Table 3. Based on ROC analysis, RAI-rev + anemic + malnourished had an AUC of 0.663, RAI-rev + anemic an AUC of 0.662, RAI-rev + malnourished an AUC of 0.639, and RAI-rev an AUC of 0.637, Figure 2B. RAI-rev + anemic + malnourished (p < 0.001) and RAI-rev + anemic (p < 0.001) superiorly predicted 30-day AEs compared to RAI-rev, Figure 2B.

3.3.3. Non-Routine Discharge

Based on multivariable analysis, RAI-rev-defined frailty (aOR: 1.11, 95% CI: 1.06–1.16) and malnourishment (aOR: 1.57, 95% CI: 1.08–2.29), Table 4, independently predicted NRD. Based on ROC analysis, RAI-rev + anemic + malnourished and RAI-rev + malnourished had identical AUCs of 0.832, and RAI-rev + anemic and RAI-rev had identical AUCs of 0.831, Figure 2C. No significant differences were noted in the models for predicting NRD, Figure 2C.

3.3.4. 30-Day Mortality

Based on multivariable analysis, 30-day mortality was independently predicted by RAI-rev-defined frailty (aOR: 1.10, 95% CI: 1.04–1.17) and malnourishment (aOR: 3.79, 95% CI: 1.24–11.60), Table 4. Based on ROC analysis, RAI-rev + anemic + malnourished had an AUC of 0.914, RAI-rev + malnourished an AUC of 0.913, and RAI-rev + anemic and RAI-rev AUC of 0.911, Figure 2D. No significant differences were noted in the models for predicting 30-day mortality, Figure 2D.

4. Discussion

In our study of 3639 ASD patients undergoing PSF, frail patients with anemia, malnutrition, or both had a higher frequency of worse outcomes, including extended LOS, NRD, readmission, and 30-day mortality. RAI-rev-defined frailty, anemia, and malnourishment independently predicted extended LOS. RAI-rev-defined frailty and anemia were risk factors for 30-day AEs while RAI-rev-defined frailty and malnutrition were associated with NRD and 30-day mortality. While the RAI-rev + anemic + malnourished and RAI-rev + anemic models were superior predictors of LOS and 30-day AEs, RAI-rev + malnutrition only outperformed RAI-rev in predicting LOS.
The impact of frailty on outcomes in postoperative ASD patients has been previously studied. In a retrospective cohort study of 1001 ASD patients, Leven et al. found that increased frailty independently predicted AEs, mortality, and unplanned reoperation [27]. In addition, in a multicenter prospective database study of 417 ASD patients, Miller et al. demonstrated increased LOS and higher rates of major complications and reoperation in frailer patients [28]. Furthermore, Baek et al., in a systematic review and meta-analysis of 474,651 degenerative spine surgery patients, demonstrated frailty to be a robust predictor of poor outcomes, including mortality, AEs, NRD, and extended LOS [29]. Similarly, our study showed RAI-rev-defined frailty to be an independent predictor of extended LOS, NRD, 30-day AEs, and mortality.
Prior investigations into anemia’s effect on ASD outcomes have demonstrated significant effects. In a retrospective cohort study of 2173 ASD patients undergoing PSF, Mo et al. found that increasing anemia severity was associated with higher odds of extended LOS, transfusion, organ space infection, and mortality [30]. Jung et al., in another retrospective database study of ASD patients, reported that patients with anemia had a higher rate of AEs, including transfusions, wound complications, and overall surgical complications [31]. In a retrospective cohort study of 473 elective posterior cervical fusion patients, Phan et al. demonstrated an increased rate of readmission, reoperation, and mortality, as well as AEs and extended LOS, in anemic patients [32]. Analogously, our results showed anemia to be an independent predictor of extended LOS and 30-day AEs.
Malnutrition has been assessed as a risk factor for adverse outcomes in ASD. In a retrospective cohort study of 303 ASD patients, Wang et al. reported a greater incidence of postoperative complications in malnourished patients [33]. Similarly, Oe et al. noted that malnutrition was an independent predictor of postoperative complications in a retrospective cohort study of 285 ASD patients [34]. In addition, a retrospective cohort study of 136 spine fusion patients by Adogwa et al. demonstrated increased complications in malnourished patients and noted malnutrition to be an independent predictor of AEs [35]. Comparatively, in our study, malnutrition independently predicted extended LOS and NRD.
While evidence demonstrating the effect of frailty, anemia, and malnutrition on postoperative outcomes is robust, few studies have evaluated the combined influence of these factors. In a retrospective cohort study of 923 digestive tract surgery patients, Li et al. noted an increased risk of adverse outcomes when combining frailty, malnutrition, and anemia [36]. In another retrospective cohort study of patients undergoing PSF, Han et al. reported that a combined frailty and malnutrition model outperformed the individual risk factors in predicting AEs [37]. Similarly, Camino-Willhuber et al., in a retrospective cohort study of 69,519 spine surgery patients, found a higher risk of complications, readmission, reoperation, and mortality among malnourished frail patients compared to nourished frail patients [38]. Our results demonstrated that models combining RAI-rev-defined frailty with malnutrition and anemia were superior predictors of postoperative outcomes in ASD patients. However, it is important to note that our model for mortality was limited by a low number of events, which may have led to model instability. Although we used AIC-guided stepwise selection to mitigate overfitting, the small event count restricted the model’s discriminatory power, as reflected by the minimal differences in AUCs across nested models. As such, findings related to mortality should be interpreted with caution.
Understanding the interplay of these risk factors will promote the development and implementation of evidence-based care pathways. For example, Enhanced Recovery After Surgery (ERAS) protocols are perioperative management guidelines that have been reported to optimize patient outcomes and decrease expenditures in spine surgery [39]. While studies assessing ERAS protocols in frail patients are limited, existing results are promising. In a retrospective cohort study of 32 frail TLIF patients, Porche et al. found that ERAS protocols shortened LOS [40]. In another retrospective cohort study of 128 frail lumbar fusion surgery patients, Cui et al. reported ERAS protocols decreased LOS and functional recovery time [41]. In terms of nutrition intervention, Chen et al. noted shorter LOS and accelerated return to normal diet with ERAS nutrition protocols [42]. Similarly, Saleh et al., in a randomized controlled trial of spine surgery patients, noted nutrition supplementation decreased complications [43]. Concerning anemia, studies have shown that preoperative treatment decreases LOS and the need for postoperative transfusions [44,45]. While these results demonstrate the utility of ERAS protocols and other focused interventions, further research is essential to better assess their efficacy in mitigating the influence of frailty, malnutrition, and anemia in spine surgery. Furthermore, considering the demonstrated efficacy of ERAS protocols, it is pertinent for future studies to evaluate the feasibility and efficacy of incorporating malnutrition and anemia screening into the existing ERAS clinical workflow.
The aforementioned studies present strong evidence for the use of perioperative risk stratification protocols and patient optimization for individual factors, like frailty, anemia, and malnutrition. Considering the results of our study demonstrating the composite model’s superior predictive capacity for 30-day AEs and extended LOS, protocols incorporating this composite model could have greater utility in identifying higher-risk patients. This composite model could be incorporated into pre-existing workflows, such as ERAS, and then run through an electronic medical database to automatically flag patients meeting high-risk criteria. Identifying high-risk patients who could benefit from optimization, such as nutrition supplementation, blood transfusions, or special frailty considerations, may enhance surgical planning, decision-making on the timing of surgery, and the utility of intraoperative measures for patients with multiple risk factors for worse outcomes. In addition, early recognition of higher-risk patients will facilitate a more individualized approach to care planning, such as at the level of pain control and anesthesia [46]. Future studies should work to assess the feasibility and cost-efficacy of implementing perioperative screening utilizing electronic medical record systems, as well as evaluate how these protocols can help to ameliorate outcomes in high-risk patients.
Our study has limitations that warrant cautious interpretation of our findings. Using the NSQIP database introduces potential coding errors, misclassification, and reporting biases. This study relied on procedural coding used in the literature to identify ASD cases, which may still be susceptible to misclassification. Inaccurate or inconsistent coding of frailty, anemia, or malnutrition could lead to misidentification of patient status, which may result in either an overestimation or underestimation of associations. Although we adjusted for numerous demographic and clinical variables in our multivariable models, residual confounding remains a concern due to unmeasured factors such as socioeconomic status (health insurance status, income), surgeon expertise, perioperative protocols, length of fusion, number of instrumented levels, and preoperative management of anemia and malnutrition. These unmeasured variables might create bias in either direction. The study’s retrospective design precludes our ability to draw causal inferences and introduces potential selection bias through non-randomized assignments. This bias may result in overestimation of effects if more severe patients are preferentially classified as frail or may lead to underestimation if protective factors are underrepresented. This adaptation of RAI-rev has not been externally validated but has been used in previous studies in the literature using NSQIP data. Furthermore, the substitution of variables may lead to misclassification which we were unable to assess. Finally, the NSQIP database only provides outcomes within a 30-day postoperative window, which may not fully capture long-term complications and mortality. Despite these limitations, to our knowledge, this is the only large-scale study assessing the combined role of frailty, malnutrition, and anemia in predicting surgical outcomes in ASD.

5. Conclusions

The results of our study reveal that RAI-rev-defined frailty has superior predictive capacity for 30-day AEs and LOS when combined with anemia and malnutrition. Future studies should further evaluate the predictive capacity of this composite model as well as assess its utility in perioperative risk stratification to inform prehabilitation measures and shared decision-making for high-risk ASD patients.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jcm14155379/s1, Table S1. List of International Classification of Disease codes used to identify spinal deformity patients; Supplementary Table S2: Score assignments for the revised Risk Analysis Index (RAI-rev); Table S3: Observations with Missing Data.

Author Contributions

Conceptualization, A.A.E.; Methodology, A.A.E. and P.S.; Formal analysis, P.S.; Investigation, A.A.E., P.S., S.D.G., J.H., E.D.L.B., S.I.K., D.S., S.-f.L.L., and D.M.S.; Data curation, A.A.E. and P.S.; Writing—original draft, A.A.E., S.D.G., and J.H.; Writing—review & editing, A.A.E., P.S., S.D.G., J.H., E.D.L.B., S.I.K., D.S., S.-f.L.L. and D.M.S.; Supervision, A.A.E., S.-f.L.L. and D.M.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Ethical review and approval were waived for this study because data was obtained from a de-identified database.

Informed Consent Statement

Patient consent was waived because data were obtained from a de-identified database.

Data Availability Statement

Data were obtained from the ACS NSQIP and are available from the authors with the permission of ACS.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
ASDAdult Spinal Deformity
ACSAmerican College of Surgeons
NSQIPNational Surgical Quality Improvement Program
PSFPosterior Spinal Fusion
ICDInternational Classification of Disease
CPTCurrent Procedural Terminology
RAI-revRevised Risk Analysis Index
ROCReceiver Operating Characteristic
LOSLength of Stay
AEAdverse Event
NRDNon-routine Discharge
OROdds Ratio
CIConfidence Interval
CHFCongestive Heart Failure
BMIBody Mass Index
ASAAmerican Society of Anesthesiologists
COPDChronic Obstructive Pulmonary Disease
SSISurgical Site Infection
PNAPneumonia
PEPulmonary Embolism
ARFAcute Renal Failure
UTIUrinary Tract Infection
MIMyocardial Infarction
DVTDeep Vein Thrombosis
RBCRed Blood Cell
MAEMinor Adverse Event
SAESevere Adverse Event
AICAkaike Information Criterion
AUCArea Under the Curve
FFrail Alone
FAFrail + Anemic
FMFrail + Malnourished
FAMFrail + Anemic + Malnourished
NFNot Frail
ERASEnhanced Recovery After Surgery

References

  1. Martin, A.B.; Hartman, M.; Washington, B.; Catlin, A.; Team, N.H.E.A. National Health Expenditures In 2023: Faster Growth As Insurance Coverage And Utilization Increased. Health Aff. 2025, 44, 12–22. [Google Scholar] [CrossRef] [PubMed]
  2. Porter, M.E. A strategy for health care reform–toward a value-based system. N. Engl. J. Med. 2009, 361, 109–112. [Google Scholar] [CrossRef]
  3. Porter, M.E. What is value in health care? N. Engl. J. Med. 2010, 363, 2477–2481. [Google Scholar] [CrossRef] [PubMed]
  4. Passias, P.G.; Jalai, C.M.; Worley, N.; Vira, S.; Hasan, S.; Horn, S.R.; Segreto, F.A.; Bortz, C.A.; White, A.P.; Gerling, M.; et al. Predictors of Hospital Length of Stay and 30-Day Readmission in Cervical Spondylotic Myelopathy Patients: An Analysis of 3057 Patients Using the ACS-NSQIP Database. World Neurosurg. 2018, 110, e450–e458. [Google Scholar] [CrossRef]
  5. Galivanche, A.R.; Gala, R.; Bagi, P.S.; Boylan, A.J.; Dussik, C.M.; Coutinho, P.D.; Grauer, J.N.; Varthi, A.G. Perioperative Outcomes in 17,947 Patients Undergoing 2-Level Anterior Cervical Discectomy and Fusion Versus 1-Level Anterior Cervical Corpectomy for Treatment of Cervical Degenerative Conditions: A Propensity Score Matched National Surgical Quality Improvement Program Analysis. Neurospine 2020, 17, 871–878. [Google Scholar] [CrossRef]
  6. De la Garza-Ramos, R.; Goodwin, C.R.; Abu-Bonsrah, N.; Jain, A.; Miller, E.K.; Neuman, B.J.; Protopsaltis, T.S.; Passias, P.G.; Sciubba, D.M. Prolonged length of stay after posterior surgery for cervical spondylotic myelopathy in patients over 65 years of age. Clin. Neurosci. 2016, 31, 137–141. [Google Scholar] [CrossRef] [PubMed]
  7. Wynkoop, E.I.; Reitenbach, M.L.; Behrens, K.M.; Tanios, M.; Kouri, A.; Khuder, S.; Risser, I.; Elgafy, H. 30- and 90-day readmission after elective spine surgery—Does postoperative inpatient medical optimization affect readmission rates: A retrospective cross-sectional study. AME Surg. J. 2024, 4, 8. [Google Scholar] [CrossRef]
  8. Kim, H.J.; Yang, J.H.; Chang, D.G.; Lenke, L.G.; Suh, S.W.; Nam, Y.; Park, S.C.; Suk, S.I. Adult Spinal Deformity: A Comprehensive Review of Current Advances and Future Directions. Asian Spine J. 2022, 16, 776–788. [Google Scholar] [CrossRef]
  9. Smith, J.S.; Shaffrey, C.I.; Bess, S.; Shamji, M.F.; Brodke, D.; Lenke, L.G.; Fehlings, M.G.; Lafage, V.; Schwab, F.; Vaccaro, A.R.; et al. Recent and Emerging Advances in Spinal Deformity. Neurosurgery 2017, 80, S70–S85. [Google Scholar] [CrossRef]
  10. Safaee, M.M.; Ames, C.P.; Smith, J.S. Epidemiology and Socioeconomic Trends in Adult Spinal Deformity Care. Neurosurgery 2020, 87, 25–32. [Google Scholar] [CrossRef]
  11. Zygourakis, C.C.; Liu, C.Y.; Keefe, M.; Moriates, C.; Ratliff, J.; Dudley, R.A.; Gonzales, R.; Mummaneni, P.V.; Ames, C.P. Analysis of National Rates, Cost, and Sources of Cost Variation in Adult Spinal Deformity. Neurosurgery 2018, 82, 378–387. [Google Scholar] [CrossRef] [PubMed]
  12. Fried, L.P.; Tangen, C.M.; Walston, J.; Newman, A.B.; Hirsch, C.; Gottdiener, J.; Seeman, T.; Tracy, R.; Kop, W.J.; Burke, G.; et al. Frailty in older adults: Evidence for a phenotype. J. Gerontol. A Biol. Sci. Med. Sci. 2001, 56, M146–M156. [Google Scholar] [CrossRef] [PubMed]
  13. Makary, M.A.; Segev, D.L.; Pronovost, P.J.; Syin, D.; Bandeen-Roche, K.; Patel, P.; Takenaga, R.; Devgan, L.; Holzmueller, C.G.; Tian, J.; et al. Frailty as a predictor of surgical outcomes in older patients. J. Am. Coll. Surg. 2010, 210, 901–908. [Google Scholar] [CrossRef]
  14. Stopa, B.M.; Robertson, F.C.; Karhade, A.V.; Chua, M.; Broekman, M.L.D.; Schwab, J.H.; Smith, T.R.; Gormley, W.B. Predicting nonroutine discharge after elective spine surgery: External validation of machine learning algorithms. J. Neurosurg. Spine 2019, 31, 742–747. [Google Scholar] [CrossRef]
  15. Yadla, S.; Ghobrial, G.M.; Campbell, P.G.; Maltenfort, M.G.; Harrop, J.S.; Ratliff, J.K.; Sharan, A.D. Identification of complications that have a significant effect on length of stay after spine surgery and predictive value of 90-day readmission rate. J. Neurosurg. Spine 2015, 23, 807–811. [Google Scholar] [CrossRef] [PubMed]
  16. Charlson, M.E.; Pompei, P.; Ales, K.L.; MacKenzie, C.R. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J. Chronic Dis. 1987, 40, 373–383. [Google Scholar] [CrossRef] [PubMed]
  17. Weaver, D.J.; Malik, A.T.; Jain, N.; Yu, E.; Kim, J.; Khan, S.N. The Modified 5-Item Frailty Index: A Concise and Useful Tool for Assessing the Impact of Frailty on Postoperative Morbidity Following Elective Posterior Lumbar Fusions. World Neurosurg. 2019, 124, e626–e632. [Google Scholar] [CrossRef]
  18. Hall, D.E.; Arya, S.; Schmid, K.K.; Blaser, C.; Carlson, M.A.; Bailey, T.L.; Purviance, G.; Bockman, T.; Lynch, T.G.; Johanning, J. Development and Initial Validation of the Risk Analysis Index for Measuring Frailty in Surgical Populations. JAMA Surg. 2017, 152, 175–182. [Google Scholar] [CrossRef]
  19. Lee, C.T.; Chen, M.Z.; Yip, C.Y.C.; Yap, E.S.; Lee, S.Y.; Merchant, R.A. Prevalence of Anemia and Its Association with Frailty, Physical Function and Cognition in Community-Dwelling Older Adults: Findings from the HOPE Study. J. Nutr. Health Aging 2021, 25, 679–687. [Google Scholar] [CrossRef]
  20. Liu, W.; Chen, S.; Jiang, F.; Zhou, C.; Tang, S. Malnutrition and Physical Frailty among Nursing Home Residents: A Cross-Sectional Study in China. J. Nutr. Health Aging 2020, 24, 500–506. [Google Scholar] [CrossRef]
  21. Stretton, B.; Booth, A.E.C.; Kovoor, J.; Gupta, A.; Edwards, S.; Hugh, T.; Maddison, J.; Talley, N.J.; Plummer, M.; Meyer, E.; et al. Impact of frailty, malnutrition and socioeconomic status on perioperative outcomes. Age Ageing 2024, 53, afae263. [Google Scholar] [CrossRef]
  22. Shah, N.V.; Kim, D.J.; Patel, N.; Beyer, G.A.; Hollern, D.A.; Wolfert, A.J.; Kim, N.; Suarez, D.E.; Monessa, D.; Zhou, P.L.; et al. The 5-factor modified frailty index (mFI-5) is predictive of 30-day postoperative complications and readmission in patients with adult spinal deformity (ASD). J. Clin. Neurosci. 2022, 104, 69–73. [Google Scholar] [CrossRef]
  23. Arya, S.; Varley, P.; Youk, A.; Borrebach, J.D.; Perez, S.; Massarweh, N.N.; Johanning, J.M.; Hall, D.E. Recalibration and External Validation of the Risk Analysis Index: A Surgical Frailty Assessment Tool. Ann. Surg. 2020, 272, 996–1005. [Google Scholar] [CrossRef]
  24. Conlon, M.; Thommen, R.; Kazim, S.F.; Dicpinigaitis, A.J.; Schmidt, M.H.; McKee, R.G.; Bowers, C.A. Risk Analysis Index and Its Recalibrated Version Predict Postoperative Outcomes Better Than 5-Factor Modified Frailty Index in Traumatic Spinal Injury. Neurospine 2022, 19, 1039–1048. [Google Scholar] [CrossRef]
  25. Maitra, S.; Mikhail, C.; Cho, S.K.; Daubs, M.D. Preoperative Maximization to Reduce Complications in Spinal Surgery. Global Spine J. 2020, 10, 45S–52S. [Google Scholar] [CrossRef]
  26. Elsamadicy, A.A.; Havlik, J.; Reeves, B.C.; Sherman, J.J.Z.; Craft, S.; Serrato, P.; Sayeed, S.; Koo, A.B.; Khalid, S.I.; Lo, S.L.; et al. Association of Malnutrition with Surgical and Hospital Outcomes after Spine Surgery for Spinal Metastases: A National Surgical Quality Improvement Program Study of 1613 Patients. J. Clin. Med. 2024, 13, 1542. [Google Scholar] [CrossRef] [PubMed]
  27. Leven, D.M.; Lee, N.J.; Kothari, P.; Steinberger, J.; Guzman, J.; Skovrlj, B.; Shin, J.I.; Caridi, J.M.; Cho, S.K. Frailty Index Is a Significant Predictor of Complications and Mortality After Surgery for Adult Spinal Deformity. Spine 2016, 41, E1394–E1401. [Google Scholar] [CrossRef] [PubMed]
  28. Miller, E.K.; Neuman, B.J.; Jain, A.; Daniels, A.H.; Ailon, T.; Sciubba, D.M.; Kebaish, K.M.; Lafage, V.; Scheer, J.K.; Smith, J.S.; et al. An assessment of frailty as a tool for risk stratification in adult spinal deformity surgery. Neurosurg. Focus. 2017, 43, E3. [Google Scholar] [CrossRef] [PubMed]
  29. Baek, W.; Park, S.Y.; Kim, Y. Impact of frailty on the outcomes of patients undergoing degenerative spine surgery: A systematic review and meta-analysis. BMC Geriatr. 2023, 23, 771. [Google Scholar] [CrossRef]
  30. Mo, K.; Ortiz-Babilonia, C.; Al Farii, H.; Raad, M.; Musharbash, F.N.; Neuman, B.J.; Kebaish, K.M. Increased Severity of Anemia Is Associated with Postoperative Complications following a Adult Spinal Deformity Surgery. World Neurosurg. 2022, 167, e541–e548. [Google Scholar] [CrossRef]
  31. Jung, A.; Kong, R.; Tracey, O.; Patel, N.; Hadid, B.; Ikwuazom, C.; Shah, N.V.; Paulino, C.B.; Monsef, J.B. Impact of iron deficiency anemia on postoperative outcomes of thoracolumbar spinal fusion (≥2-level) on patients with adult spinal deformity with minimum two-year follow-up surveillance. Spine J. 2022, 22, S60–S61. [Google Scholar] [CrossRef]
  32. Phan, K.; Dunn, A.E.; Kim, J.S.; Capua, J.D.; Somani, S.; Kothari, P.; Lee, N.J.; Xu, J.; Dowdell, J.E.; Cho, S.K. Impact of Preoperative Anemia on Outcomes in Adults Undergoing Elective Posterior Cervical Fusion. Global Spine J. 2017, 7, 787–793. [Google Scholar] [CrossRef]
  33. Wang, J.; Oe, S.; Yamato, Y.; Hasegawa, T.; Yoshida, G.; Banno, T.; Arima, H.; Mihara, Y.; Ide, K.; Watanabe, Y.; et al. Preoperative Malnutrition-Associated Spinal Malalignment with Patient-Reported Outcome Measures in Adult Spinal Deformity Surgery: A 2-Year Follow-Up Study. Spine Surg. Relat. Res. 2023, 7, 74–82. [Google Scholar] [CrossRef]
  34. Oe, S.; Yamato, Y.; Hasegawa, T.; Yoshida, G.; Kobayashi, S.; Yasuda, T.; Banno, T.; Arima, H.; Mihara, Y.; Ushirozako, H.; et al. Association between a prognostic nutritional index less than 50 and the risk of medical complications after adult spinal deformity surgery. J. Neurosurg. Spine 2020, 33, 219–224. [Google Scholar] [CrossRef]
  35. Adogwa, O.; Martin, J.R.; Huang, K.; Verla, T.; Fatemi, P.; Thompson, P.; Cheng, J.; Kuchibhatla, M.; Lad, S.P.; Bagley, C.A.; et al. Preoperative serum albumin level as a predictor of postoperative complication after spine fusion. Spine 2014, 39, 1513–1519. [Google Scholar] [CrossRef]
  36. Li, C.Q.; Zhang, C.; Yu, F.; Li, X.Y.; Wang, D.X. The composite risk index based on frailty predicts postoperative complications in older patients recovering from elective digestive tract surgery: A retrospective cohort study. BMC Anesthesiol. 2022, 22, 7. [Google Scholar] [CrossRef] [PubMed]
  37. Han, D.; Wang, P.; Wang, S.K.; Cui, P.; Lu, S.B. Frailty and malnutrition as predictors of major complications following posterior thoracolumbar fusion in elderly patients: A retrospective cohort study. Spine J. 2024, 25, 679–687. [Google Scholar] [CrossRef] [PubMed]
  38. Camino-Willhuber, G.; Tani, S.; Schonnagel, L.; Caffard, T.; Haffer, H.; Chiapparelli, E.; Sarin, M.; Shue, J.; Soffin, E.M.; Zelenty, W.D.; et al. Association of Frailty and Preoperative Hypoalbuminemia with the Risk of Complications, Readmission, and Mortality After Spine Surgery. World Neurosurg. 2023, 174, e152–e158. [Google Scholar] [CrossRef]
  39. Pennington, Z.; Cottrill, E.; Lubelski, D.; Ehresman, J.; Theodore, N.; Sciubba, D.M. Systematic review and meta-analysis of the clinical utility of Enhanced Recovery After Surgery pathways in adult spine surgery. J. Neurosurg. Spine 2021, 34, 325–347. [Google Scholar] [CrossRef]
  40. Porche, K.; Yan, S.; Mohamed, B.; Garvan, C.; Samra, R.; Melnick, K.; Vaziri, S.; Seubert, C.; Decker, M.; Polifka, A.; et al. Enhanced recovery after surgery (ERAS) improves return of physiological function in frail patients undergoing one- to two-level TLIFs: An observational retrospective cohort study. Spine J. 2022, 22, 1513–1522. [Google Scholar] [CrossRef]
  41. Cui, P.; Wang, S.; Wang, P.; Yang, L.; Kong, C.; Lu, S. Comparison of perioperative outcomes in frail patients following multilevel lumbar fusion surgery with and without the implementation of the enhanced recovery after surgery protocol. Front. Surg. 2022, 9, 997657. [Google Scholar] [CrossRef] [PubMed]
  42. Chen, C.H.; Hsieh, H.-M. The Introduction of ERAS Nutritional Care Can Improve the Quality of Postoperative Care. Curr. Dev. Nutr. 2022, 6, 738. [Google Scholar] [CrossRef]
  43. Saleh, H.; Williamson, T.K.; Passias, P.G. Perioperative Nutritional Supplementation Decreases Wound Healing Complications Following Elective Lumbar Spine Surgery: A Randomized Controlled Trial. Spine 2023, 48, 376–383. [Google Scholar] [CrossRef] [PubMed]
  44. Vitale, M.G.; Stazzone, E.J.; Gelijns, A.C.; Moskowitz, A.J.; Roye, D.P. The effectiveness of preoperative erythropoietin in averting allogenic blood transfusion among children undergoing scoliosis surgery. J. Pediatr. Orthop. B 1998, 7, 203–209. [Google Scholar] [CrossRef]
  45. Cahill, C.M.; Alhasson, B.; Blumberg, N.; Melvin, A.; Knight, P.; Gloff, M.; Robinson, R.; Akwaa, F.; Refaai, M.A. Preoperative anemia management program reduces blood transfusion in elective cardiac surgical patients, improving outcomes and decreasing hospital length of stay. Transfusion 2021, 61, 2629–2636. [Google Scholar] [CrossRef]
  46. Villafañe, J.H.; Valdes, K.; Pedersini, P.; Berjano, P. Osteoarthritis: A call for research on central pain mechanism and personalized prevention strategies. Clin. Rheumatol. 2019, 38, 583–584. [Google Scholar] [CrossRef]
Figure 1. Flow diagram for inclusion and exclusion criteria.
Figure 1. Flow diagram for inclusion and exclusion criteria.
Jcm 14 05379 g001
Figure 2. Comparing model performance for predicting (A) extended hospital length of stay (LOS), (B) 30-day adverse events (AEs), (C) non-routine discharge (NRD), and (D) 30-day mortality using area under the receiver operating characteristic (AUROC) curves. (A) p-values: RAI-rev + anemic vs. RAI rev: p < 0.001; RAI-rev + malnourished vs. RAI-rev: p < 0.001; RAI-rev + anemic + malnourished: p < 0.001; Models adjusted for sex, ASA classification, dependent functional status, and bleeding disorder. (B) p values: RAI-rev + anemic vs. RAI rev: p < 0.001; RAI-rev + malnourished vs. RAI-rev: p = 0.403; RAI-rev + anemic + malnourished: p < 0.001; Models adjusted for age, sex, race, ASA classification, diabetes mellitus, disseminated cancer, and smoking. (C) p values: RAI-rev + anemic vs. RAI rev: p = 0.724; RAI-rev + malnourished vs. RAI-rev: p = 0.609; RAI-rev + anemic + malnourished: p = 0.522; Models adjusted for age, sex, BMI, hypertension, diabetes mellitus, COPD, disseminated cancer, SAE, hospital length of stay, and operation time. (D) p values: RAI-rev + anemic vs. RAI rev: p = 0.996; RAI-rev + malnourished vs. RAI-rev: p = 0.810; RAI-rev + anemic + malnourished: p = 0.760; Models adjusted for age, electrolyte abnormality, smoking, preoperative steroids, bleeding disorder, MAE, and SAE.
Figure 2. Comparing model performance for predicting (A) extended hospital length of stay (LOS), (B) 30-day adverse events (AEs), (C) non-routine discharge (NRD), and (D) 30-day mortality using area under the receiver operating characteristic (AUROC) curves. (A) p-values: RAI-rev + anemic vs. RAI rev: p < 0.001; RAI-rev + malnourished vs. RAI-rev: p < 0.001; RAI-rev + anemic + malnourished: p < 0.001; Models adjusted for sex, ASA classification, dependent functional status, and bleeding disorder. (B) p values: RAI-rev + anemic vs. RAI rev: p < 0.001; RAI-rev + malnourished vs. RAI-rev: p = 0.403; RAI-rev + anemic + malnourished: p < 0.001; Models adjusted for age, sex, race, ASA classification, diabetes mellitus, disseminated cancer, and smoking. (C) p values: RAI-rev + anemic vs. RAI rev: p = 0.724; RAI-rev + malnourished vs. RAI-rev: p = 0.609; RAI-rev + anemic + malnourished: p = 0.522; Models adjusted for age, sex, BMI, hypertension, diabetes mellitus, COPD, disseminated cancer, SAE, hospital length of stay, and operation time. (D) p values: RAI-rev + anemic vs. RAI rev: p = 0.996; RAI-rev + malnourished vs. RAI-rev: p = 0.810; RAI-rev + anemic + malnourished: p = 0.760; Models adjusted for age, electrolyte abnormality, smoking, preoperative steroids, bleeding disorder, MAE, and SAE.
Jcm 14 05379 g002
Table 1. Patient demographics and comorbidities by health condition combination.
Table 1. Patient demographics and comorbidities by health condition combination.
VariablesFrail Alone
(n = 460)
Frail +
Anemic
(n = 266)
Frail + Malnourished
(n = 37)
Frail +
Anemic + Malnourished
(n = 121)
Not Frail
(n = 2755)
p-Value
Age (years),
mean (SD)
70.94 (7.69)69.85 (8.05)73.08 (9.24)70.47 (8.75)45.50 (17.48)<0.001 *
Female, n (%)245 (53.3)82 (30.8)21 (56.8)57 (47.1)1992 (72.3)<0.001 *
Race/Ethnicity,
n (%)
0.005 *
 NHW373 (85.9)186 (76.9)18 (64.3)85 (76.6)1866 (80.6)
 NHB27 (6.2)26 (10.7)3 (10.7)13 (11.7)242 (10.4)
 Hispanic16 (3.7)21 (8.7)5 (17.9)6 (5.4)127 (5.5)
 Other18 (4.1)9 (3.7)2 (7.1)7 (6.3)81 (3.5)
BMI (kg/m2),
mean (SD)
29.33 (6.02)29.07 (6.84)29.17 (6.70)28.10 (6.43)27.49 (7.25)<0.001 *
ASA, n (%) <0.001 *
 12 (0.4)0 (0.0)0 (0.0)0 (0.0)186 (6.8)
 2118 (25.7)38 (14.3)7 (18.9)16 (13.2)1190 (43.3)
 3328 (71.3)199 (75.1)24 (64.9)75 (62.0)1302 (47.3)
 ≥412 (2.6)28 (10.6)6 (16.2)30 (24.8)72 (2.6)
Hypertension,
n (%)
324 (70.4)193 (72.6)23 (62.2)86 (71.1)879 (31.9)<0.001 *
Diabetes mellitus,
n (%)
79 (17.2)66 (24.8)12 (32.4)35 (28.9)244 (8.9)<0.001 *
COPD, n (%)28 (6.1)19 (7.1)3 (8.1)9 (7.4)82 (3.0)<0.001 *
CHF, n (%)9 (2.0)9 (3.4)2 (5.4)2 (1.7)5 (0.2)<0.001 *
Dependent functional status,
n (%)
33 (7.2)31 (11.7)3 (8.1)34 (28.1)95 (3.5)<0.001 *
Disseminated cancer, n (%)14 (3.0)25 (9.4)7 (18.9)12 (9.9)0 (0.0)<0.001 *
Electrolyte abnormality,
n (%)
28 (6.1)25 (9.5)5 (13.5)26 (21.5)143 (5.9)<0.001 *
Smoking, n (%)33 (7.2)46 (17.3)7 (18.9)18 (14.9)540 (19.6)<0.001 *
Pre-operative steroids, n (%)24 (5.2)19 (7.1)2 (5.4)21 (17.4)111 (4.0)<0.001 *
Bleeding disorder,
n (%)
10 (2.2)17 (6.4)2 (5.4)4 (3.3)43 (1.6)<0.001 *
SD: standard deviation; NHW: non-Hispanic White; NHB: non-Hispanic Black; BMI: body mass index; ASA: American Society of Anesthesiologists classification; COPD: chronic obstructive pulmonary disease; CHF: congestive heart failure; * Statistically significant, unadjusted p-value.
Table 2. Rates of postoperative outcomes by health condition combination.
Table 2. Rates of postoperative outcomes by health condition combination.
VariablesFrail Alone
(n = 460)
Frail + Anemic
(n = 266)
Frail + Malnourished
(n = 37)
Frail +
Anemic + Malnourished
(n = 121)
Not Frail
(n = 2755)
p-Value
Any AE, n (%)213 (46.3)161 (60.5)16 (43.2)75 (62.0)1259 (45.7)<0.001 *
Surgical AEs,
n (%)
 Superficial SSI5 (1.1)8 (3.0)0 (0.0)2 (1.7)37 (1.3)0.214
 Deep SSI4 (0.9)2 (0.8)1 (2.7)1 (0.8)25 (0.9)0.842
 Organ space SSI4 (0.9)5 (1.9)0 (0.0)3 (2.5)33 (1.2)0.488
 Wound dehiscence2 (0.4)3 (1.1)0 (0.0)1 (0.8)23 (0.8)0.830
Medical AEs,
n (%)
 PNA13 (2.8)14 (5.3)2 (5.4)5 (4.1)66 (2.4)0.051
 Reintubation3 (0.7)10 (3.8)0 (0.0)8 (6.6)36 (1.3)<0.001 *
 Ventilator requirement8 (1.7)12 (4.5)0 (0.0)10 (8.3)40 (1.5)<0.001 *
 PE16 (3.5)11 (4.1)1 (2.7)2 (1.7)36 (1.3)0.001 *
 Renal insufficiency5 (1.2)6 (2.5)0 (0.0)0 (0.0)12 (0.5)0.004 *
 ARF1 (0.2)2 (0.8)0 (0.0)1 (0.8)3 (0.1)0.093
 UTI22 (4.8)15 (5.6)2 (5.4)4 (3.3)62 (2.3)0.001 *
 Cardiac arrest or MI3 (0.7)2 (0.8)0 (0.0)2 (1.7)10 (0.4)0.260
 DVT7 (1.5)9 (3.4)0 (0.0)2 (1.7)30 (1.1)0.032 *
C. diff colitis0 (0.0)3 (1.5)0 (0.0)1 (1.2)6 (0.3)0.039 *
 Systemic sepsis12 (2.6)8 (3.0)1 (2.7)4 (3.3)53 (1.9)0.576
 Septic shock2 (0.4)2 (0.8)1 (2.7)5 (4.1)9 (0.3)<0.001 *
 Postoperative RBC transfusion174 (37.8)145 (54.5)12 (32.4)64 (52.9)1143 (41.5)<0.001 *
AE severity, n (%)
 MAE42 (9.1)38 (14.3)4 (10.8)11 (9.1)166 (6.0)<0.001 *
 SAE197 (42.8)156 (58.6)13 (35.1)70 (57.9)1221 (44.3)<0.001 *
Hospital length of stay (days), mean (SD)6.41 (5.09)8.92 (7.58)12.61 (11.30)13.49 (12.72)6.50 (6.24)<0.001 *
Total operation time (hours), mean (SD)5.49 (2.69)5.41 (2.76)4.64 (2.67)4.80 (2.29)5.51 (2.62)0.017 *
Readmission, n (%)40 (8.7)28 (10.6)3 (8.1)14 (11.6)171 (6.3)0.010 *
Reoperation, n (%)26 (5.7)16 (6.0)4 (10.8)8 (6.6)143 (5.2)0.568
Non-routine discharge,
n (%)
210 (46.2)154 (58.6)28 (77.8)79 (65.8)587 (21.4)<0.001 *
Mortality, n (%)3 (0.7)4 (1.5)1 (2.7)9 (7.4)9 (0.3)<0.001 *
AE: adverse event; SSI: surgical site infection; PNA: pneumonia; PE: pulmonary embolism; ARF: acute renal failure; UTI: urinary tract infection; MI: myocardial infarction; DVT: deep vein thrombosis; C. diff: Clostridioides difficile; RBC: red blood cell; MAE: minor adverse event; SAE: severe adverse event; SD: standard deviation; * Statistically significant, unadjusted p-value.
Table 3. Multivariable logistic regression model on the odds of extended hospital length of stay and 30-day adverse events.
Table 3. Multivariable logistic regression model on the odds of extended hospital length of stay and 30-day adverse events.
Adjusted ORLower Limit
95% CI
Upper Limit
95% CI
p-Value
Extended Hospital Length of Stay
RAI-rev1.031.011.04<0.001 *
Anemic1.841.452.35<0.001 *
Malnourished2.341.693.24<0.001 *
Female1.391.101.750.006 *
ASA
 1–2REFREFREFREF
 ≥31.901.472.47<0.001 *
Dependent functional status2.051.393.03<0.001 *
Bleeding disorder1.680.913.120.098
30-Day Adverse Events
RAI-rev1.071.041.11<0.001 *
Anemic2.091.662.61<0.001 *
Malnourished0.930.671.280.646
Age0.980.960.990.001 *
Female2.171.722.75<0.001 *
Race/Ethnicity
 NHWREFREFREFREF
 NHB0.660.480.910.012 *
 Hispanic0.670.441.020.061
 Other1.250.742.090.400
ASA
 1–2REFREFREFREF
 ≥ 31.541.241.91<0.001 *
Diabetes mellitus0.810.621.060.123
Disseminated cancer0.280.120.620.002 *
Smoking0.600.470.77<0.001 *
OR: odds ratio; CI: confidence interval; RAI-rev: revised risk analysis index; ASA: American Society of Anesthesiologists classification; NHW: non-Hispanic White; NHB: non-Hispanic Black; * Statistically significant, unadjusted p-value.
Table 4. Multivariable logistic regression model on the odds of nonroutine discharge and 30-day mortality.
Table 4. Multivariable logistic regression model on the odds of nonroutine discharge and 30-day mortality.
Adjusted ORLower Limit
95% CI
Upper Limit
95% CI
p-Value
Nonroutine Discharge
RAI-rev1.111.061.16<0.001 *
Anemic1.250.971.620.090
Malnourished1.571.082.290.018 *
Age1.031.011.05<0.001 *
Female1.451.091.920.010 *
BMI1.021.001.040.034 *
Hypertension0.620.480.80<0.001 *
Diabetes mellitus1.361.001.850.053
COPD1.410.892.250.145
Disseminated cancer0.190.070.49<0.001 *
SAE1.651.282.14<0.001 *
LOS (days)1.091.071.12<0.001 *
Operation time (hours)1.141.081.19<0.001 *
30-Day Mortality
RAI-rev1.101.041.170.001 *
Anemic1.040.343.170.951
Malnourished3.791.2411.600.020 *
Age0.970.941.000.081
Electrolyte abnormality2.540.837.780.102
Smoking0.250.032.070.199
Pre-operative steroids5.742.1015.69<0.001 *
Bleeding disorder3.190.7513.600.117
MAE6.442.4716.80<0.001 *
SAE2.940.939.370.068
OR: odds ratio; CI: confidence interval; RAI-rev: revised risk analysis index; BMI: body mass index; COPD: chronic obstructive pulmonary disease; SAE: severe adverse event; LOS: hospital length of stay; MAE: minor adverse event; * Statistically significant, unadjusted p-value.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Elsamadicy, A.A.; Serrato, P.; Ghanekar, S.D.; Hansen, J.; Brown, E.D.L.; Khalid, S.I.; Schneider, D.; Lo, S.-f.L.; Sciubba, D.M. Composite RAI, Malnutrition, and Anemia Model Superiorly Predicts 30-Day Morbidity and Mortality After Surgery for Adult Spinal Deformity. J. Clin. Med. 2025, 14, 5379. https://doi.org/10.3390/jcm14155379

AMA Style

Elsamadicy AA, Serrato P, Ghanekar SD, Hansen J, Brown EDL, Khalid SI, Schneider D, Lo S-fL, Sciubba DM. Composite RAI, Malnutrition, and Anemia Model Superiorly Predicts 30-Day Morbidity and Mortality After Surgery for Adult Spinal Deformity. Journal of Clinical Medicine. 2025; 14(15):5379. https://doi.org/10.3390/jcm14155379

Chicago/Turabian Style

Elsamadicy, Aladine A., Paul Serrato, Shaila D. Ghanekar, Justice Hansen, Ethan D. L. Brown, Syed I. Khalid, Daniel Schneider, Sheng-fu Larry Lo, and Daniel M. Sciubba. 2025. "Composite RAI, Malnutrition, and Anemia Model Superiorly Predicts 30-Day Morbidity and Mortality After Surgery for Adult Spinal Deformity" Journal of Clinical Medicine 14, no. 15: 5379. https://doi.org/10.3390/jcm14155379

APA Style

Elsamadicy, A. A., Serrato, P., Ghanekar, S. D., Hansen, J., Brown, E. D. L., Khalid, S. I., Schneider, D., Lo, S.-f. L., & Sciubba, D. M. (2025). Composite RAI, Malnutrition, and Anemia Model Superiorly Predicts 30-Day Morbidity and Mortality After Surgery for Adult Spinal Deformity. Journal of Clinical Medicine, 14(15), 5379. https://doi.org/10.3390/jcm14155379

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop