Readmission rate after epilepsy surgery: A systematic review and meta-analysis

Objective Today, quality of care has become a central issue with regard to redesigning health care delivery. For improvement of quality of care, measurement and analysis of outcomes is essential. the aim of this study is to evaluate the 30 day readmission rate of epilepsy surgery. Material and Methods A systematic literature review was performed in April 2020. We reviewed MEDLINE/PubMed, Cochrane Library, and ClinicalTrials.gov for citation or ongoing trials from January 2010 to March 2020. The search criteria were limited to human studies published in English language. The Medical Subject Headings terms used for the search in PubMed were ‘epilepsy surgery’ ,’readmission’, ’reoperation ’ and ‘neurological surgery’. We used random-effects meta-analyses to estimate pooled risk ratios and 95% condence intervals for early readmission Results In mean Elixhauser


Results
In the 7trials reporting readmission rates, the overall prevalence of readmission and re-operation within 30 days was 10% and 9%. The most common cause of re-admission was seizure(31.6%). The overall mean Elixhauser index among readmitted patients after epilepsy surgery was 8.05.

Conclusion
Minimizing factors that contribute to readmission in various patient populations and procedures becomes important for patient care, resource utilization, and physician reimbursement.understanding about a readmission rate of 10% can be useful to health systems designing quality improvement efforts.Awareness of the reasons for readmission is important for patient counseling and surgical decision making.In epilepsy surgery, although our study did not show a signi cant difference but ,readmissions might be reduced with optimization of pre-existing comorbidities Introduction: today, quality of care has become a central issue with regard to redesigning health care delivery. For improvement of quality of care, measurement and analysis of outcomes is essential. One outcome that is of interest in terms of measuring quality of care and hospital performance is the 30-day readmission rate (1). Jencks et al. recently reported a pro le of readmitted patientsamong the Medicare population (2). In this study, nearly 20% of Medicare bene ciaries who had been discharged from a hospital were readmitted within 30 days, and the estimated cost to Medicare of unplanned readmissions in a single year was $17.4 billion. Concerns that some readmissions might be preventable have in uenced policy changes to simultaneously improve quality of care and lower costs (3). Hospitals at which Medicare riskadjusted readmission rates are greater than the national average are at risk for losing reimbursements.
Recent standards established by the Centers for Medicare & Medicaid Services (CMS) have initiated programs to reduce the economic burden of hospital readmission by penalizing hospitals with higher than expected readmission rates (4). Readmissions contribute to higher operating costs for health-care systems as seen with unplanned 30-day readmissions costing Medicare $17.4 billion in the year of 2004 (5). Epilepsy surgery is highly effective in appropriately seected candidates with drug-resistant epilepsy (6). A nationwide study examining 30-day readmissions after surgical resection for drug-resistant epilepsy noted an 11.5% read-mission rate. Other studies using the Agency for Healthcare Research Quality (AHRQ) State Inpatient Database (SID) reported readmission rates of up to 14% for seizures after resective epilepsy surgery (7). These policies addressing all-cause readmissions necessitate a better understanding of the factors that in uence readmission. Although extensive literature about readmission attributed to particular conditions, especially congestive heart failure, is available, there remains a paucity of research on readmissions within neurosurgical practices (8). A recent, single-institution study measuring neurosurgical outcomes after more than 5000 procedures reported rates of major complications but did not include readmissions. The remaining information is limited to cohorts undergoing spine surgery, for whom reported 30-day readmission rates range from 4.2-9.1%. In all studies that we found, the most common reason for readmission was infection (9)(10). To elucidate the rate, reason, and predictors of readmissions after seizure interventions at a tertiary/quaternary neurosurgical service, Such analyses are critical for de ning the problem so that the area to target for quality improvement within neurosurgery can be identi ed.the aim of this study is to evaluate the 30 day readmission rate of epilepsy surgery.

Eligibility Criteria
Eligible studies were randomized trials reported in English, since 2010, that assessed the rate of readmissions in 30 days after epilepsy surgery.The intervention had to focus its efforts on the hospital-tohome transition, permit patients across arms to have otherwise similar inpatient experiences, and be generalizable to contexts beyond a single patient diagnosis. Adult patients had to be admitted from the community to an inpatient ward for at least 24 hours with a medical or surgical cause. Studies including obstetric or psychiatric admissions or only including discharges to skilled nursing or rehabilitation facilities were excluded.

Data Sources And Search Strategy
A systematic literature review was performed in April 2020. We reviewed MEDLINE/PubMed, Cochrane Library, and ClinicalTrials.gov for citation or ongoing trials from January 2010 to March 2020. The search criteria were limited to human studies published in English language. The Medical Subject Headings terms used for the search in PubMed were 'epilepsy surgery' ,'readmission', 'reoperation ' and 'neurological surgery' .

Data Extraction Data extraction was conducted independently according to the Preferred Reporting Items for Systematic
Reviews and Meta Analyses. Two authors selected studies following the previously described inclusion criteria. After this rst selection, if needed, another author was consulted to achieve a shared decision. A prede ned protocol for data extraction was used to retrieve data of each study, including rst author name, year of publication, journal of publication, sample size, study design, demographic pro le,readmission rate,re-operation rate, and signi cant variable at uni and multivariate analysis.

Risk Of Bias
Two raters (M.S and FP.) worked independently and in duplicate to determine the extent to which each trial was at risk of bias using a standardized form based on the Cochrane Colaboration's tool.The assessment considered the quality of the randomization sequence generation, allocation concealment, blinding of outcome assessors, the potential for missing outcomes (ie, likelihood of missing readmissions to other hospitals), and the proportion of patients lost to follow-up. For missing outcomes, "high risk of bias" was assigned when the readmissions data came from internal health system records only. To assess for publication bias, we examined a funnel plot for asymmetry and conducted asymmetry regression according to Sterne and Egger and determined the associated P value.

Data Synthesis
We used random-effects meta-analyses to estimate pooled risk ratios and 95% con dence intervals for early readmission. We tested for heterogeneity of effect on this outcome using the Cochran Q χ2 test and estimated between-trial inconsistency not due to chance using the I2 statistic.To explore the effects of patient, intervention, and out come characteristics on the impact of measured intervention effectiveness, we conducted planned subgroup analyses, testing variables 1 at a time. Patient characteristics tested were age, gender, and hospital ward (general medical or other). Ad hoc variables tested were year of publication and type of outcome reported (ie, unplanned readmissions).

Study Selection
Our initial database search generated 327 reports (Fig. 1)Through abstract and title screening, 319 reports were identi ed for full-text review. During full-text screening 7 were selected for inclusion and 53 were set aside for author contact prior to making a decision. Table 2 describes the included trials. Many were single-center trials taking place in academic medical centers, enrolling few patients and 30-day readmissions. Most interventions tested took place in both the inpatient and outpatient settings. Most studies were at low risk of bias .The most common methodological limitation of these trials was the lack of a reliable method for dealing with missing data.
Meta-analysis of Median annual household income among readmitted patients after epilepsy surgery : The total prevalence of readmitted patients with being in the 0-25th percentile, 26th -50th percentile and Meta-analysis of prevalence of surgical and medical complications in readmitted patients after epilepsy surgery: The total prevalence of surgical and medical complications among readmitted patients after epilepsy surgery were 9.4%(95% CI,8.6%,10.2% ,I 2 :99.5% ), and 5%(95% CI,4.3%,5.7%), respectively.

Meta-analysis of the mean length of stay (LOS) among readmitted patients after epilepsy surgery:
The overall mean length of stay among readmitted patients after epilepsy surgery was 2.95 days.
Meta-analysis of the mean Elixhauser index among readmitted patients after epilepsy surgery: The overall mean Elixhauser index among readmitted patients after epilepsy surgery was 8.05.
Meta-analysis of the prevalence of seizure as the most common cause of readmission among patients after epilepsy surgery: The overall prevalence of seizure among readmitted patients after epilepsy surgery was 31.6%(95% CI,30.2%,32.9% ,I 2 :99.6% ) Meta-regression nding based on the mean of age and frequency of readmission: The studies' meta-regression was according to the association between frequency of readmission and the mean age of patients.There was no statistically signi cant linear trend in univariate meta-regression to explain effect size variation by mean of age of study with coe cient = 0.14 (95% CI − 2.17, 2.46), P = 0.88 (Fig. 3a).
Meta-regression nding based on the male to female ratio of study and prevalence of readmission: The overall rate of readmissions based on the female to male ratio of the studies is showed in Figure. 4b,the prevalence of readmission was lower in studies with higher male to female ratio. There was statistically signi cant linear trend in univariate meta-regression to explain effect size variation by male to female ratio of study with coe cient = 0.47 (95% CI 0.03, 162 0.91), P = 0.03. (Fig. 3b) Meta-regression nding based on the mean of age and frequency of readmission: The studies' meta-regression was according to the association between frequency of readmission and the publication year of included studies.There was no statistically signi cant linear trend in univariate meta-regression to explain effect size variation by mean of age of study with coe cient = 0.14 (95% CI − 2.17, 2.46), P = 0.88 (Fig. 3c).
Publication Bias: Funnel plot in Figure (3d) shows no indication of publication bias. It is shows in funnel plot symmetrically. Circles' size shows the weight of studies (bigger circles shows more sample and smaller circles shows fewer sample).

Discussion:
It is well known that epilepsy surgery success rates range from 50-80%. (11,12) Hence, it is not surprising that patients might require readmission for additional monitoring or management of persistent seizures. There is a growing body of literature about 30-day readmission rates after neurosurgical interventions.
Single neurosurgical centers have found overall readmission rates nearing 10%, with most predictors of readmissions seen secondary to postoperative care complications (13)(14)(15). However, within this body of literature, those related to surgery in persons with epilepsy remain limited.The principle aim of this systematic review and meta-analysis was to determine the 30-day readmission rate in epilepsy surgery, which we found to be 10%.all included studies were from USA and reported 30-day readmission using NSQIP database. Predictors of readmission has been reported to be male sex and initial admission via the emergency department(16), Medicare insurance ,lowest quartile of income, depression, hemispherectomy, postoperative complications, and small bed size hospital (17).in our study due to the lack of data in the included studies we could only analize some of the effective factors such as Payer, Mean Elixhauser index, Median annual household income, Lenghth of stay, Surgical complications, Medical complications, Hospital bed size and Median household income.our study showed no signi cant relationship between any of these factors and the readmission rate.the most common cause of readmission was seizure .deppression has been reported to be a predictor of worse postsurgical seizure outcome, speci cally, less likely to be seizure free (18,19). the mean Elixhauser index was 8.05 and the mean length of stay was 2.95 days. Rumalla et al in their study evaluated the 90 day readmission too.they reported a 16.5% 90 day readmission. They also indicated that Predictors of 90-d readmission, in addition to Medicare payer status and depression, included increasing number of comorbidities and medical complications (20). The most common medical complications likely resulted from mechanical ventilation and immobility or deep venous thrombosis (DVT). Therefore, additional attention is deserved for preventing these medical complications, ie, DVT prophylaxis, early ambulation, and incentive spirometry (21). Moghavem et al in their study stated that More evidence is required to identify the factors that may lead to improved outcomes, particularly modi able factors. Some of our data suggest that many readmissions identi ed in our study are potentially preventable complications such as (postoperative infections, urinary tract infections,pneumonia, and pulmonary emboli). Thirty-day readmission metrics are now commonly studied as they can represent an indicator of quality of care and cost-effective management. Monetary incentives are distributed by healthcare regulations targeting quality improvement in inpatient management that is also cost-saving. Our results suggest that, particularly in Medicare recipients who tend to have more comorbid conditions, careful postoperative discharge planning that involves allied health services after neurosurgery and follow-up soon after discharge to minimize seizures or any signs of post-op complications may be important to enhance outcomes and reduce the risk of readmission and its associated consequences in epilepsy.

Limitation:
First ,It is possible that additional causes of readmission listed in secondary diagnosis elds were missed. Since second,the follow-up time was limited in some of the included studies. Our study may also be understating the risk of readmission for epilepsy surgery, as the data of included studies do not capture patients who die outside the hospital after their procedure or seek additional treatment in another state.

Conclusion:
Minimizing factors that contribute to readmission in various patient populations and procedures becomes important for patient care, resource utilization, and physician reimbursement.understanding about a readmission rate of 10% can be useful to health systems designing quality improvement efforts.Awareness of the reasons for readmission is important for patient counseling and surgical decision making.In epilepsy surgery, although our study did not show a signi cant difference but ,readmissions might be reduced with optimization of pre-existing comorbidities. Author contribution: ZM participated in Conception and design of the study, library searches and assembling relevant literature, critical review of the paper, supervising writing of the paper, Database management.FP participated in Data collection, library searches and assembling relevant literature, writing the paper, and critical review of the paper.MS participated in Data collection, library searches and assembling relevant literature, writing the paper, analysis of the data and critical review of the paper. All authors read and approved the nal manuscript.  Meta-analysis of the readmission and re-operation prevalence Figure 3 Meta-regression nding based on the mean of age(a),publication year(b),male to female ratio(c) and frequency of readmission. Begg's funnel plot for publication bias (d)