https://orcid.org/0000-0002-1843-7732
Figures
Abstract
Objective
The purpose of this study was to present the first comprehensive analysis of perioperative quality indicators; length of hospital stay; readmission; reoperation; pre-, intra, and postoperative events; and mortality in a diverse neurosurgical patient cohort in Europe.
Methods
Electronic medical records of all patients who were admitted to our institution between January 1 and December 31 of 2020, and underwent an index neurosurgical operation (n = 1142) were retrospectively reviewed.
Results
The median length of hospital stay at the index admission and readmission was 8 days (range: 1–242 days) and 5 days (range: 0–94 days), respectively. Of the 1142 patients, 22.9% (n = 262) had an extended length of hospital stay of ≥14 days. The all-cause 7-, 15-, 30-, 60-, and 90-day readmission rates were 3.9% (n = 44), 5.7% (n = 65), 8.8% (n = 100), 12.3% (n = 141), and 16.5% (n = 188), respectively. The main reason for unplanned readmission was deterioration of medical and/or neurological condition. The all-cause 7-, 15-, 30-, 60-, and 90-day reoperation rates were 11.1% (n = 127), 13.8% (n = 158), 16.5% (n = 189), 18.7% (n = 213), and 19.4% (n = 221), respectively. Unplanned reoperations were due primarily to hydrocephalus. The rate of preoperative events was 1.1% (n = 13), one-third of which were associated with infection. The rate of intraoperative events was 11.0% (n = 126), of which 98 (64.47%) were surgical, 37 (24.34%) were anesthesiologic, and 17 (11.18%) were associated with technical equipment. The rate of postoperative events was 9.5% (n = 109). The most common postoperative event was malfunction, disconnection, or dislocation of an implanted device (n = 24, 17.91%). The mortality rates within 7, 15, 30, 60, and 90 days after the index operation were 0.9% (n = 10), 1.8% (n = 21), 2.5% (n = 29), 3.4% (n = 39), and 4.7% (n = 54), respectively. Several patient characteristics and perioperative factors were significantly associated with outcome parameters.
Citation: Chavush E, Rössler K, Dorfer C (2024) Perioperative quality indicators among neurosurgery patients: A retrospective cohort study of 1142 cases at a tertiary center. PLoS ONE 19(2): e0297167. https://doi.org/10.1371/journal.pone.0297167
Editor: Dean Chou, Columbia University Vagelos College of Physicians and Surgeons, UNITED STATES
Received: August 10, 2023; Accepted: December 30, 2023; Published: February 6, 2024
Copyright: © 2024 Chavush et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: AIDS, Acquired immune deficiency syndrome; ASA, American Society of Anesthesiology; BMI, Body mass index; ER, Emergency room; HIV, Human immunodeficiency virus; ICD-10, International Classification of Diseases 10th Version; ICU, Intensive care unit; LOS, Length of hospital stay; NSAID, Non-steroidal anti-inflammatory drug; WHO, World Health Organization
Introduction
In recent years, growing emphasis has been placed on value-based healthcare rather than volume-based healthcare. Advancing the efficiency of healthcare delivery by improving patient outcomes relative to costs not only benefits patients but also increases the economic sustainability of healthcare systems [1]. Accordingly, parameters assessing patient outcomes, such as the length of hospital stay (LOS), readmission rate, reoperation rate, adverse event rate, and mortality rate, are increasingly used as quality metrics [2–8].
The average LOS varies greatly among countries and healthcare systems, and this variation cannot be explained by medical reasons or necessity [9, 10]. As reported by the Organisation for Economic Co-operation and Development, the average LOS in Austria is longer than that in Anglo-American and Scandinavian countries, but shorter than that in Germany [11]. Decreasing the LOS in German-speaking countries is important, because research has suggested that each day spent in the hospital confers additional risks on patients regardless of their primary diagnoses [12]. Hospital-acquired conditions such as air embolism, blood incompatibility, catheter-associated infections, poor glycemic control, and surgical site infections continue to markedly threaten patient safety [13].
Hospital readmission rates can indicate substandard care and unfavorable patient outcomes [4, 14, 15]. As a quality parameter, they indicate the effect of hospital care on patient condition after discharge [16]. Rehospitalization occurs frequently and may continue to be life-threatening [17]. Because inpatient acute care is the most expensive type of health service, hospital readmission results in substantial healthcare costs [16]. Research has suggested that a sizable proportion of readmissions may potentially be avoidable [14].
Reoperation rates are increasing being used as another quality indicator, because of their wide applicability and straightforward determination [18]. According to a study in five European hospitals, 1 minute of operating room time costs approximately 9.45 € in a conventional surgical theater [19]. In hybrid operating rooms, which are also frequently used for neurosurgical procedures, the average cost is more than twice as high, at 19.88 € [19]. Given that neurosurgical operations generally have longer durations than operations in other specialties, unplanned returns to the operating room can greatly increase the overall costs of patient care [20].
Adverse events after neurosurgical procedures are not rare [7]. These events not only endanger patients’ health status, but are also associated with extended LOS, readmission, and reoperation, thereby resulting in substantial healthcare costs [5, 7, 21]. The incidence of adverse events after surgical operations is expected to rise even further, as a consequence of aging of the population and increased burden of illness worldwide [21]. In previous studies, prolonged length of preoperative hospital stay has been identified as a modifiable risk factor for various postoperative adverse events [22–25]. However, an overview of in-hospital complications between hospital admission and index operation is currently not available. Moreover, the overall incidence rate of intraoperative complications among neurosurgery patients has not been investigated in the literature to date. Earlier work in this area was either procedure specific or focused on only pediatric patients. Moreover, no report to date has described the predictive factors associated with intraoperative complications in neurosurgery [26–30].
Therefore, the aim of this study was to present what is, to our knowledge, the first analysis of perioperative quality indicators in a diverse neurosurgical patient cohort in a European healthcare system.
Materials and methods
Design and setting
This study was a retrospective cohort analysis of patients at the Department of Neurosurgery of the Vienna General Hospital, Medical University of Vienna. Approval was granted by the ethics committee of the Medical University of Vienna (number 1360/2021), in accordance with the issued Guidelines for Good Scientific Practice. The requirement for informed patient consent was waived by the ethics committee. The Vienna General Hospital is a tertiary referral center, and the largest hospital in Austria and fifth largest hospital in Europe. It has a capacity of 1732 beds [31], and more than 61000 inpatient admissions and more than 1.1 million outpatient visits per year [32].
Electronic medical records of all patients (n = 1142) who were admitted to our institution between January 1 and December 31 of 2020, and underwent an index neurosurgical operation were retrospectively reviewed. Data collection was performed through the patient information management system (AKIM) of Vienna General Hospital by the first author between June 1, 2021, and July 20, 2022. Only the first author had access to information that could be used to identify individual participants during or after data collection.
Outcome variables
The primary outcome measures of interest were the LOS at the index admission and readmission; rates of all-cause and unplanned readmission within 7, 15, 30, 60, and 90 days after discharge from the index admission; causes of readmission; International Classification of Diseases 10th Version (ICD-10) diagnosis codes at readmission; rates of all-cause and unplanned reoperation within 7, 15, 30, 60, and 90 days after the index operation; ICD-10 diagnosis codes at reoperation; procedures performed at reoperation; causes of reoperation; pre-, intra-, and postoperative events and their rates; time until adverse events; and mortality rates within 7, 15, 30, 60, and 90 days after the index operation.
Definitions of outcomes
In this study, the LOS spanned from hospital admission to discharge. Extended LOS was defined as a hospital stay exceeding the 75th percentile for the overall cohort (≥14 days). Index hospital admission refers to the first hospital admission of patients to the Department of Neurosurgery of the Vienna General Hospital, in which neurosurgical operation was performed. The index operation was defined as the first neurosurgical operation performed at the Department of Neurosurgery of the Vienna General Hospital. Readmission refers to return to the Vienna General Hospital for any cause after prior discharge. Reoperation describes a repeat neurosurgical procedure after the index surgery. Preoperative events were defined as any adverse event that occurred in the hospital between the index admission and the index operation. Intraoperative events describe any adverse events occurring in the operating theater during the index surgery, as described by the operating surgeon. Postoperative events were defined as any abnormal or unexpected neurosurgical events occurring within 90 days after the index surgery, and contributing to prolonged hospitalization, readmission, or reoperation. Detailed definitions of postoperative events are provided in Table 10 in the S1 Appendix.
Statistical analysis
SPSS® statistics software, version 29 (IBM Corporation), was used for statistical analysis. Descriptive statistics, including means, medians, rates, ranges, percentages, and standard deviations, was used to describe the collected patient datasets. Univariate analyses between categorical variables were performed with Pearson’s χ2 test, Fisher’s exact test, the Fisher-Freeman-Halton exact test, or logistic regression, as appropriate. A Monte Carlo simulation was applied, if necessary, with a confidence level of 99% and 10000 samples. Continuous variables were compared via Mann-Whitney U test or independent-samples Kruskal-Wallis test with Dunn-Bonferroni post hoc analysis. Multivariate logistic regression models were built with an entry criterion of p<0.05 to further investigate the independent predictors of extended LOS, unplanned readmission, unplanned reoperation, perioperative events, and mortality. The Hosmer–Lemeshow test was used to determine the calibration of the models. A p-value less than 0.05 was defined as statistically significant. All reported p-values are two-sided.
Results
The baseline characteristics of the 1142 patients in the study population are presented in Table 1. Fig 1 in the S1 Appendix shows the distribution of study participants in cohorts.
The evaluated operative metrics and their outcomes are summarized in Table 2. The preoperative laboratory values of the patients and their outcomes are shown in Table 1 in the S1 Appendix. The assessments of additional patient characteristics and clinical parameters are reported in Table 2 in the S1 Appendix. A subgroup analysis of pediatric patients in various age groups is shown in Table 3 in the S1 Appendix.
In the study cohort, the most common diagnoses at the index hospital admission were primary brain tumor (n = 334, 29.2%), unruptured cerebral aneurysm (n = 75, 6.6%), secondary brain tumor (n = 70, 6.1%), epilepsy (n = 68, 6%), and subdural hemorrhage (n = 61, 5.3%) (Table 3). Among the 368 patients treated for a primary cranial or spinal tumor, meningioma (n = 97, 26.4%), glioblastoma (n = 51, 13.9%), vestibular schwannoma (n = 48, 13%), pituitary adenoma (n = 42, 11.4%), and pilocytic astrocytoma (n = 13, 3.5%) were the most frequent tumor histologies (Table 4 in the S1 Appendix). WHO grade I (n = 142, 38.6%) and WHO grade IV (n = 58, 15.8%) tumors represented most of the primary neoplasms in the patient cohort (Table 5 in the S1 Appendix). The main origins of the secondary tumors (n = 72) were lung cancer (n = 38, 47.2%), breast cancer (n = 8, 11.1%), B-cell lymphoma (n = 5, 6.9%), and melanoma (n = 4, 5.6%) (Table 6 in the S1 Appendix).
The most frequently performed procedures at the index operation were removal of an intracerebral mass in the cerebral hemispheres (n = 174, 15.2%), catheter angiography of the head and neck (n = 77, 6.7%), removal of extracerebral tumors at the skull base of the brain (n = 63, 5.5%), decompression of lumbar nerve roots (n = 63, 5.5%), transsphenoidal resection of an adenoma (n = 51, 4.5%), placement of an external ventricular drain through burr hole trepanation (n = 46, 4%), evacuation of a subdural hematoma through burr hole trepanation (n = 45, 3.9%), removal of a cerebellar tumor (n = 41, 3.6%), resection of a vestibular schwannoma (n = 40, 3.5%), and resection of epileptogenic foci (n = 31, 2.7%). The outcomes for the most frequent index surgeries are summarized in Table 4.
Length of hospital stay
The overall LOS of the entire patient cohort at the index hospitalization ranged from a 1 to 242 days. On average, the patients spent a median of 8 days at the hospital during the index admission. The median LOS until the index operation was 1 day (range: 0–50 days), whereas the median LOS after the index operation was 7 days (range: 0–242 days). Patients with an extended LOS remained in the hospital for a median of 21.5 days (range: 14–242 days). In contrast, patients without prolonged hospitalization stayed in the hospital for a median of 7 days (range: 1–13 days). The cohort of patients with hydrocephalus had the longest median LOS (median 14 days, range 3–111 days), whereas the shortest median LOS was observed among patients undergoing spinal neurosurgery (median 6 days, range 1–74 days) (Table 1). The median length of the index hospitalization for the most frequent tumor histologies was 9 days for meningioma (range: 3–127 days), 13 days for glioblastoma (range: 2–87 days), 9 days for vestibular schwannoma (5–53 days), 8 days for pituitary adenoma (range: 5–30 days), and 10 days for pilocytic astrocytoma (range: 6–21 days). Patients with WHO grade I, II, or III tumors spent a median of 9 days in the hospital during the index stay, whereas patients with WHO grade IV tumors remained in the hospital for a median of 13 days.
Extended length of hospital stay
Of the 1142 patients in the study cohort, 262 (22.9%) had an extended LOS, i.e., hospitalization for ≥14 days. The prevalence of extended LOS was highest among neonates (83.3%, n = 10), followed by toddlers 12 months to <2 years of age (37.5%, n = 3), and young children 2 to <6 years of age (36.7%, n = 11).
Patient characteristics predictive of an extended LOS in univariate analyses were high Charlson Comorbidity Index score (OR 1.89, p = 0.010); presence of hemiplegia or paraplegia (OR 4.69, p = 0.003), depression (OR 2.44, p = 0.012), or diabetes without end-organ damage (OR 1.89, p = 0.021); high number of preoperative medications (OR 2.09, p = 0.001); preoperative use of anticoagulants (OR 2.04, p = 0.002), steroids (OR 1.54, p = 0.044), angiotensin II receptor blockers (OR 1.67, p = 0.015), calcium channel blockers (OR 1.59, p = 0.034), insulin (OR 2.69, p = 0.048), hypoglycemic oral drugs (OR 2.42, p = 0.003), or β-blockers (OR 1.62, p = 0.019); and high ASA grade (OR 10.40, p<0.001) An assessment of social parameters indicated that patients who were unable to independently perform daily activities in the home environment were relatively more likely to have an extended LOS at the index hospitalization (OR 0.24, p<0.001).
Multivariate analysis indicated that being in the cohort of patients with hydrocephalus (OR 7.69, 95% CI 1.75–33.85, p = 0.007), preoperative albumin of <35 g/L (OR 4.05, 95% CI 1.38–11.92, p = 0.011), preoperative prothrombin time >125% (Owren) (OR 2.53, 95% CI 1.04–6.17, p = 0.041), presence of preoperative depression (OR 3.34, 95% CI 1.04–10.69, p = 0.043), transfer from another department (OR 15.79, 95% CI 5.52–45.17, p<0.001), transfer from another hospital (OR 2.76, 95% CI 1.36–5.61, p = 0.005), two neurosurgical operations during the index hospital stay (OR 5.66, 95% CI 2.88–11.11, p<0.001), three or more neurosurgical operations during the index hospital stay (OR 11.07, 95% CI 3.39–36.13, p<0.001), intensive care unit stay (OR 2.41, 95% CI 1.30–4.49, p = 0.005), readmission to the intensive care unit during the index hospital stay (OR 5.14, 95% CI 1.07–24.01, p = 0.041), intraoperative event at the index operation (OR 2.77, 95% CI 1.47–5.23, p = 0.002), postoperative event within 90 days after the index operation (OR 5.52, 95% CI 2.76–11.04, p<0.001), and transfer to another department after surgical treatment (OR 54.02, 95% CI 16.71–174.66, p<0.001) were independently associated with greater likelihood of an extended hospital stay.
Readmissions
Among the 1142 participants, the all-cause 7-, 15-, 30-, 60-, and 90-day readmission rates were 3.9% (n = 44), 5.7% (n = 65), 8.8% (n = 100), 12.3% (n = 141), and 16.5% (n = 188), respectively. The median duration until readmission was 34 days (range: 0–90 days) after discharge from the index hospitalization and 48 days (range: 4–196 days) after the index operation (Fig 1). Of the 247 readmissions, 127 (51.4%) were planned, and 120 (48.6%) were unplanned. The 7-, 15-, 30-, 60-, and 90-day unplanned readmission rates were 2.8% (n = 32), 3.7% (n = 42), 5.3% (n = 61), 7.0% (n = 80), and 8.5% (n = 97), respectively. Patients underwent neurosurgical operation in approximately one-third of readmissions (n = 82, 33.2%). More than three-quarters of readmissions (n = 188, 76.1%) were first-time rehospitalizations; the maximum number of readmission episodes within 90 days after discharge from the index hospital admission was 7 (n = 1, 0.4% of readmission). The median LOS was 4 days (range: 0–56 days) for planned readmission and 5.5 days (range: 0–94 days) for unplanned readmission.
Unplanned readmission was due primarily to medical and/or neurological deterioration (n = 26, 21.67%), wound complications (n = 12, 10%), and non-central nervous system infection (n = 11, 9.17%) (Table 5).
The most frequently documented diagnoses at unplanned readmission were “C71.9 Malignant neoplasm: Brain, unspecified” (n = 13, 10.83%), “T81.3 Disruption of operation wound, not elsewhere classified” (n = 11, 9.17%), and “C79.3 Secondary malignant neoplasm of brain and cerebral meninges” (n = 5, 4.17%).
According to the logistic regression model, a preoperative sodium level of <135 mmol/L (OR 4.35, 95% CI 1.48–12.80, p = 0.008), prior deep vein thrombosis (OR 5.05, 95% CI 1.17–21.87, p = 0.030), presence of psychosis preoperatively (OR 16.74, 95% CI 1.82–153.90, p = 0.013), being in the pediatric neurosurgery cohort (OR 2.49, 95% CI 1.01–6.12, p = 0.047), transfer from another department during the index hospitalization (OR 6.88, 95% CI 2.48–19.10, p<0.001), transfer from another hospital during the index hospitalization (OR 3.45, 95% CI 1.73–6.87, p<0.001), and the occurrence of a postoperative event within 90 days after the index surgery (OR 17.76, 95% CI 9.57–32.96, p<0.001) were identified as independent factors associated with an elevated risk of unplanned readmission.
Reoperations
The 1142 patients in the study cohort underwent a total of 352 reoperations within 90 days after the index surgery. The all-cause 7-, 15-, 30-, 60-, and 90-day reoperation rates were 11.1% (n = 127), 13.8% (n = 158), 16.5% (n = 189), 18.7% (n = 213), and 19.4% (n = 221), respectively. Approximately 56% of the reoperations were unplanned (n = 200, 56.8%). Among the study participants, the rates of unplanned reoperation within 7, 15, 30, 60, and 90 days after the index surgery were 4.6% (n = 53), 6.7% (n = 77), 9.3% (n = 106), 10.6% (n = 121), and 11% (n = 126), respectively. More than three-quarters of reoperations were performed during the index hospitalization (n = 268, 76.1%), and patients required readmission to the Vienna General Hospital in a total of 84 repeated interventions (23.9%). The number of reoperations per patient ranged from 1 to 14. The median durations until planned and unplanned reoperation were 4 days (range: 0–83 days) and 13 days (range: 0–86 days), respectively (Fig 2 in the S1 Appendix).
Unplanned reoperation was most frequently due to hydrocephalus (n = 50, 25%) (pediatric and adult study populations), postoperative hemorrhage (n = 21, 10.5%), or external ventricular drainage-associated complications (n = 18, 9%) (Table 7 in the S1 Appendix). The most common ICD-10 diagnoses at unplanned reoperation were “G91.9 Hydrocephalus, unspecified” (n = 27, 13.5%), “I60.9 Subarachnoid hemorrhage, unspecified” (n = 24, 12%), and “I62.0 Nontraumatic subdural hemorrhage” (n = 11, 5.50%). In unplanned reoperations, the most frequently performed surgeries were “placement of a ventricular shunt” (n = 42, 21%), followed by “placement of an external ventricular drain” (n = 41, 20.50%), and “other operation—neurocranium and dura” (n = 18, 9%).
Multivariate analyses indicated that age ≥90 years (OR 70. 84, 95% CI 2.66–1884.25, p = 0.011), urgent index operation (OR 4.77, 95% CI 1.50–15.17, p = 0.008), emergent index operation (OR 8.61, 95% CI 1.91–38.82. p = 0.005), LOS of 15–21 days at the index hospitalization (OR 13.43, 95% CI 1.63–110.49, p = 0.016), LOS of >21 days at the index hospitalization (OR 62.83, 95% CI 7.48–527.50, p<0.001), transfer to a rehabilitation center after the index hospitalization (OR 119.57, 95% CI 1.39–10318.09, p = 0.035), discharge against medical advice after the index hospitalization (OR 12.31, 95% CI 1.45–104.37, p = 0.021), and the occurrence of a postoperative event within 90 days after the index surgery (OR 157.86, 95% CI 56.79–438.76, p<0.001) were statistically associated with unplanned reoperation.
Perioperative events
In the study cohort, 13 patients experienced a total of 15 preoperative events. The overall rate of preoperative events was 1.1%. The duration until the preoperative events was known in 8 of 15 cases, and amounted to a median of 0 days (range: 0–11 days). Approximately three-quarters of the preoperative events occurred in the general ward (n = 11, 73.33%), whereas the rest were reported during intensive care unit or intermediate care unit stays (n = 4, 26.67%). The rate of intraoperative events in the patient cohort was 11% (n = 126). The total number of intraoperative events was 152. Of these, 98 (64.47%) were surgical, 37 (24.34%) were anesthesiologic, and 17 (11.18%) were technical equipment-related complications (Table 9 in the S1 Appendix). No patients died in the operating theater. The rate of postoperative events was 9.5% (n = 109). The most common postoperative events after elective procedures were pituitary surgery associated events (n = 14) and wound complications (n = 13), whereas the malfunction, disconnection, or dislocation of an implanted device was the most frequently described event after non-elective procedures (n = 13) (Table 11 in the S1 Appendix). A time-to-event analysis indicating the average time until the occurrence of postoperative events is shown in Fig 2. All events requiring immediate attention, such as hemorrhage or infarcts, occurred within 4 days after surgery. Factors significantly associated with perioperative events in multivariate logistic regression models are summarized in Table 12 in the S1 Appendix.
Mortality
The mortality rates within 7, 15, 30, 60, and 90 days after the index operation were 0.9% (n = 10), 1.8% (n = 21), 2.5% (n = 29), 3.4% (n = 39), and 4.7% (n = 54), respectively. The proportion of patients who died within 90 days after the index surgery was significantly higher among patients who underwent an emergency operation (n = 17, 19.1%) (OR 10.11, p<0.001) than an elective operation (n = 18, 2.3%). The median time to death was 25 days (range: 1–90 days). The number of female and male patients who died was equal (n = 27, 50%). More than half of the deaths occurred after patient discharge from the index hospitalization (n = 29, 53.7%), approximately one-third of the patients died in the intensive care unit or intermediate care unit (n = 18, 33.33%), and 12.96% of the patients died in the general ward (n = 7). the intraoperative mortality rate was 0% (n = 0). Kaplan-Meier curves depicting the time to death between patients with or without extended LOS, 90-day readmission, or 90-day reoperation are presented in Figs 3–5 in the S1 Appendix, respectively.
Discussion
Regular measurement and analysis of outcome parameters is crucial in evaluating the quality of healthcare systems and enabling necessary steps toward improvement.
In our study, the median LOS at the index hospitalization was 8 days, which was longer than the reported neurosurgical hospitalizations in Anglo-American countries [33, 34]. Approximately one-fifth of patients had an extended LOS of at least 14 days. Our findings highlighted that, despite an observed slow decline in LOS in the nine neurosurgical departments in Austria over the past two decades [35], further initiatives are necessary to avoid hospital days beyond medical readiness [36].
The main reason for the discrepancy in the LOS between European countries such as Austria and countries such as the United States of America lies in historically deep-seated cultural differences regarding when patient discharge is considered safe by physicians and patients, in the absence of any data supporting that longer hospital stays increase safety. Acquiring evidence-based data is the only way to change the perceptions and traditions that have resulted in patients remaining in hospitals until their stiches have been removed, as reflected by the median hospital stay of 8 days. Our study provides evidence that patients without identified risk factors—most importantly neurological compromise, diminished albumin, elevated prothrombin time at admission, and intraoperative events—could be discharged earlier in the health care environment of Austria. All early complications requiring immediate attention, such as postoperative hemorrhage or infarcts, occurred within 4 days after surgery. All late complications seen after discharge were readily identified in the outpatient setting, and were associated primarily with wound complications and infections.
Furthermore, to decrease the LOS among neurosurgical patients in Austria, special attention must be paid to patients after transphenoidal pituitary surgery, given that a substantial number of events associated with hyponatremia may potentially be life-threatening. Similarly, diminished levels of potassium in patients with a cardiac comorbidity should be normalized before admission. Steroid use in patients admitted for brain tumor surgery should be critically assessed for reduction or termination, if possible, given that unnecessary steroid prescription is associated with an increased complication rate and extended hospital stay [37–40].
In addition, a concomitant decrease in preventable unplanned readmission, thus avoiding unnecessary costs, can be accomplished only if the risk factors for readmission have been identified. Herein, we identified that a preoperative sodium level < 135 mm/L, prior deep vein thrombosis, the presence of psychosis, and transfer from another department or hospital were independent risk factors for unplanned readmission among neurosurgical patients. With regard to unplanned reoperation, we performed a detailed analysis and identified that age ≥90 years, urgent or emergent index operation, extended LOS, and transfer to a rehabilitation center after index hospitalization, among other factors, were statistically associated with unplanned reoperation.
All the above-identified factors must be considered when regulatory measures to decrease the LOS in neurosurgery are developed in health care policies. In Austria, increased resource allocation for outpatient care would be necessary to provide a safe environment for neurosurgical patients when the LOS is further decreased [38]. Similarly, for optimization of preoperative outpatient care, we identified a preoperative in-hospital event rate of 1.1%. Most in-hospital adverse events occurring before the index surgery were associated with infection. Our results emphasize that patients are at risk of complications even before they undergo surgery; therefore unnecessary inpatient days before surgery, as well as surgical delays after hospital admission, should be avoided. We also identified that several abnormalities in preoperative laboratory values were associated with preoperative events. These findings suggest that patients who are predisposed to complications could be identified, and their clinical condition could be optimized before hospital admission for an elective neurosurgical operation.
Finally, our study provides a definition of postoperative events that could be used for future comparative analyses of neurosurgical patient populations, by defining a postoperative adverse events as any abnormal or unexpected neurosurgical event occurring within 90 days after the index surgery, and contributing to a prolonged hospitalization, readmission, or reoperation [7, 41–44]. In our cohort, the proportion of patients who experienced a postoperative event, according to this definition, was 8.4% (n = 96).
Strengths and limitations
The major limitation of our study is its single-center design, which limits the generalizability of our results to other centers in Austria. Although a comparison with the three other large academic centers in Austria might have increased the reliability of the findings, given that patients with the most complex cases requiring comprehensive care are often referred to those centers, the selection bias toward less complex surgeries in the five regional hospitals might have been too pronounced to enable a detailed analysis. However, by taking the unspecific measure of the LOS into account, clear parallels can be seen, thereby allowing for an indirect extrapolation [35]. Furthermore, because Austria is a small country with no more than 9 million inhabitants, any major differences in patient care across the nine neurosurgical centers, which could potentially have greatly influenced the outcome parameters analyzed herein, would be expected to be small, but nonetheless must be critically addressed. Furthermore, the study period’s spanning January through December of 2020, in the midst of the COVID-19 pandemic, must be critically considered. In contrast to many other neurosurgical departments worldwide, patient care at our institution was not affected by the COVID-19 pandemic to an extent that could have made our study population not representative in this specific regard [45]. Among the multifactorial reasons for the lack of influence of the pandemic at our institution, the most important reason was that our neurosurgical department is situated in a separate building from the main hospital, where all the other departments are located. Therefore, neither the neurosurgical ICU nor our general ward’s bed capacity was affected by any restrictions during the study period. Indeed, the overall case load was slightly greater than that in previous years, thus indicating that the any limitations due to the study period itself might have been counterintuitively small.
Simultaneously, our analysis encompassed the entire spectrum of neurosurgical diseases and procedures across all age groups. Furthermore, because our analyses were not based solely on administrative datasets with coded information, we were able to collect extensive and precise information on patient characteristics, perioperative factors, and outcomes. Nonetheless, the descriptions of intraoperative events in this study were restricted to the reports of the operating surgeons.
Conclusion
This study identified risk factors for prolonged LOS, readmission, and reoperation in a diverse neurosurgical patient cohort in Europe. The findings highlight LOS among neurosurgical patients in Austria could feasibly be decreased if outpatient care were tailored to the specific needs of this vulnerable patient population.
References
- 1. Porter ME. What is value in health care? The New England journal of medicine. 2010;363(26):2477–81. pmid:21142528
- 2. Thomas JW, Guire KE, Horvat GG. Is patient length of stay related to quality of care? Hospital & health services administration. 1997;42(4):489–507. pmid:10174462
- 3. Tsai TC, Joynt KE, Orav EJ, Gawande AA, Jha AK. Variation in surgical-readmission rates and quality of hospital care. The New England journal of medicine. 2013;369(12):1134–42. pmid:24047062
- 4. Ashton CM, Del Junco DJ, Souchek J, Wray NP, Mansyur CL. The association between the quality of inpatient care and early readmission: a meta-analysis of the evidence. Medical care. 1997;35(10):1044–59. pmid:9338530
- 5. Mukerji N, Jenkins A, Nicholson C, Mitchell P. Unplanned reoperation rates in pediatric neurosurgery: a single center experience and proposed use as a quality indicator. Journal of neurosurgery Pediatrics. 2012;9(6):665–9. pmid:22656260
- 6. Suero Molina E, Schildmacher C, Doods J, Freistühler M, Hellwig SJ, Stummer W, et al. The rise of quality indicators in neurosurgery: 30-day unplanned reoperation rate evaluated in 3760 patients-a single-center experience. Acta neurochirurgica. 2020;162(1):147–56. pmid:31802277
- 7. Ziewacz JE, McGirt MJ, Chewning SJ Jr. Adverse events in neurosurgery and their relationship to quality improvement. Neurosurgery clinics of North America. 2015;26(2):157–65, vii. pmid:25771271
- 8. Wahba AJ, Cromwell DA, Hutchinson PJ, Mathew RK, Phillips N. Mortality as an indicator of quality of neurosurgical care in England: a retrospective cohort study. BMJ open. 2022;12(11):e067409. pmid:36332948
- 9. Collins TC, Daley J, Henderson WH, Khuri SF. Risk factors for prolonged length of stay after major elective surgery. Annals of surgery. 1999;230(2):251–9. pmid:10450740
- 10. Patil PG, Turner DA, Pietrobon R. National trends in surgical procedures for degenerative cervical spine disease: 1990–2000. Neurosurgery. 2005;57(4):753–8; discussion -8. pmid:16239888
- 11.
OECD. Length of hospital stay [Internet]. [Cited 2023 May 17]. Available from: https://data.oecd.org/healthcare/length-of-hospital-stay.htm.
- 12. Krumholz HM. Post-hospital syndrome—an acquired, transient condition of generalized risk. The New England journal of medicine. 2013;368(2):100–2. pmid:23301730
- 13.
Centers for Medicare & Medicaid Services. Hospital-Acquired Conditions [Internet]. [Cited 2023 May 17]. Available from: https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/HospitalAcqCond/Hospital-Acquired_Conditions.
- 14. McIlvennan CK, Eapen ZJ, Allen LA. Hospital readmissions reduction program. Circulation. 2015;131(20):1796–803. pmid:25986448
- 15. Zuckerman RB, Sheingold SH, Orav EJ, Ruhter J, Epstein AM. Readmissions, Observation, and the Hospital Readmissions Reduction Program. The New England journal of medicine. 2016;374(16):1543–51. pmid:26910198
- 16. Westert GP, Lagoe RJ, Keskimäki I, Leyland A, Murphy M. An international study of hospital readmissions and related utilization in Europe and the USA. Health policy (Amsterdam, Netherlands). 2002;61(3):269–78. pmid:12098520
- 17. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. The New England journal of medicine. 2009;360(14):1418–28. pmid:19339721
- 18. Kroon HM, Breslau PJ, Lardenoye JW. Can the incidence of unplanned reoperations be used as an indicator of quality of care in surgery? American journal of medical quality: the official journal of the American College of Medical Quality. 2007;22(3):198–202. pmid:17485561
- 19. Patel S, Lindenberg M, Rovers MM, van Harten WH, Ruers TJM, Poot L, et al. Understanding the Costs of Surgery: A Bottom-Up Cost Analysis of Both a Hybrid Operating Room and Conventional Operating Room. International journal of health policy and management. 2022;11(3):299–307. pmid:32729284
- 20. Costa ADS Jr. Assessment of operative times of multiple surgical specialties in a public university hospital. Einstein (Sao Paulo, Brazil). 2017;15(2):200–5. pmid:28767919
- 21. Ludbrook GL. The Hidden Pandemic: the Cost of Postoperative Complications. Current anesthesiology reports. 2022;12(1):1–9. pmid:34744518
- 22. Cediel EG, Boerwinkle VL, Ramon JF, Arias D, De la Hoz-Valle JA, Mercado JD, et al. Length of preoperative hospital stay is the dominating risk factor for surgical site infection in neurosurgery: A cohort data-driven analysis. Surgical neurology international. 2022;13:80. pmid:35399909
- 23. deFreitas DJ, Kasirajan K, Ricotta JJ 2nd, Veeraswamy RK, Corriere MA. Preoperative inpatient hospitalization and risk of perioperative infection following elective vascular procedures. Annals of vascular surgery. 2012;26(1):46–54. pmid:22079458
- 24. Bartels SA, Gardenbroek TJ, Bos L, Ponsioen CY, D’Haens GR, Tanis PJ, et al. Prolonged preoperative hospital stay is a risk factor for complications after emergency colectomy for severe colitis. Colorectal disease: the official journal of the Association of Coloproctology of Great Britain and Ireland. 2013;15(11):1392–8. pmid:23810064
- 25. Bueno Cavanillas A, Rodrìguez-Contreras R, Delgado Rodriguez M, Moreno Abril O, López Gigosos R, Guillen Solvas J, et al. Preoperative stay as a risk factor for nosocomial infection. European journal of epidemiology. 1991;7(6):670–6. pmid:1783060
- 26. Aleksic V, Radulovic D, Milakovic B, Nagulic M, Vucovic D, Antunovic V, et al. A retrospective analysis of anesthesiologic complications in pediatric neurosurgery. Paediatric anaesthesia. 2009;19(9):879–86. pmid:19627531
- 27. van Lindert EJ, Delye H, Leonardo J. Prospective review of a single center’s general pediatric neurosurgical intraoperative and postoperative complication rates. Journal of neurosurgery Pediatrics. 2014;13(1):107–13. pmid:24236448
- 28. van Lindert EJ, Arts S, Blok LM, Hendriks MP, Tielens L, van Bilsen M, et al. Intraoperative complications in pediatric neurosurgery: review of 1807 cases. Journal of neurosurgery Pediatrics. 2016;18(3):363–71. pmid:27231823
- 29. Fridriksson S, Säveland H, Jakobsson KE, Edner G, Zygmunt S, Brandt L, et al. Intraoperative complications in aneurysm surgery: a prospective national study. Journal of neurosurgery. 2002;96(3):515–22. pmid:11883836
- 30. Arts S, Delye H, van Lindert EJ. Intraoperative and postoperative complications in the surgical treatment of craniosynostosis: minimally invasive versus open surgical procedures. Journal of neurosurgery Pediatrics. 2018;21(2):112–8. pmid:29171801
- 31.
The Vienna General Hospital. Number of beds classified according to type (2022) [Internet]. [Cited 2023 June 8]. Available from: https://www.akhwien.at/default.aspx?pid=871.
- 32.
The Vienna General Hospital. Key figures for the medical units (2021) [Internet]. [Cited 2023 June 8]. Available from: https://www.akhwien.at/default.aspx?pid=866.
- 33. Hersh DS, Smith LGF, Jones TL, Fraser BD, Kumar R, Vaughn B, et al. Predictors of an Extended Length of Stay following an Elective Craniotomy in Children and Young Adults. Pediatric neurosurgery. 2020;55(5):259–67. pmid:33099552
- 34. Dasenbrock HH, Liu KX, Devine CA, Chavakula V, Smith TR, Gormley WB, et al. Length of hospital stay after craniotomy for tumor: a National Surgical Quality Improvement Program analysis. Neurosurgical focus. 2015;39(6):E12. pmid:26621410
- 35.
Chavush E. Assessment of Perioperative Quality Indicators Among Neurosurgery Patients—A Comprehensive Cross-Sectional Analysis of Over One Thousand Cases [Internet] [Thesis]. [Medical University of Vienna]; 2023 [cited 2023 Oct 19]. Available from: https://ubsearch.meduniwien.ac.at/permalink/f/cg0si2/UMW_alma5148977390003344
- 36.
Federal Ministry of Social Affairs, Health, Care and Consumer Protection [Internet]. [Cited 2023 May 19]. Available from: https://www.sozialministerium.at/Themen/Gesundheit/Gesundheitssystem/Gesundheitssystem-und-Qualitaetssicherung/Qualitaetsentwicklung-im-oesterreichischen-Gesundheitswesen.html.
- 37. Dietrich J, Rao K, Pastorino S, Kesari S. Corticosteroids in brain cancer patients: benefits and pitfalls. Expert review of clinical pharmacology. 2011;4(2):233–42. pmid:21666852
- 38. Kotsarini C, Griffiths PD, Wilkinson ID, Hoggard N. A systematic review of the literature on the effects of dexamethasone on the brain from in vivo human-based studies: implications for physiological brain imaging of patients with intracranial tumors. Neurosurgery. 2010;67(6):1799–815; discussion 815. pmid:21107211
- 39. Petrelli F, De Stefani A, Ghidini A, Bruschieri L, Riboldi V, Dottorini L, et al. Steroids use and survival in patients with glioblastoma multiforme: a pooled analysis. Journal of neurology. 2021;268(2):440–7. pmid:32002651
- 40. Lee EQ, Wen PY. Corticosteroids for peritumoral edema: time to overcome our addiction? Neuro-oncology. 2016;18(9):1191–2. pmid:27530501
- 41. Houkin K, Baba T, Minamida Y, Nonaka T, Koyanagi I, Iiboshi S. Quantitative analysis of adverse events in neurosurgery. Neurosurgery. 2009;65(3):587–94; discussion 94. pmid:19687705
- 42. Rolston JD, Han SJ, Lau CY, Berger MS, Parsa AT. Frequency and predictors of complications in neurological surgery: national trends from 2006 to 2011. Journal of neurosurgery. 2014;120(3):736–45. pmid:24266542
- 43. Wong JM, Panchmatia JR, Ziewacz JE, Bader AM, Dunn IF, Laws ER, et al. Patterns in neurosurgical adverse events: intracranial neoplasm surgery. Neurosurgical focus. 2012;33(5):E16. pmid:23116096
- 44. Fugate JE. Complications of Neurosurgery. Continuum (Minneapolis, Minn). 2015;21(5 Neurocritical Care):1425–44. pmid:26426239
- 45. Mischkulnig M, Hopp B, Wadiura LI, Khalaveh F, Kiesel B, Rössler K, et al. Treatment of high-grade glioma patients during the COVID-19 pandemic: Impact on overall survival, tumor size and delay of treatment. PLOS ONE. 2023;18(6):e0287993. pmid:37390078