OBES SURG DOI 10.1007/s11695-015-1763-y
A Comparison of Bariatric Surgery in Hospitals With and Without ICU: a Linked Data Cohort Study David J. R. Morgan 1
Kwok M. Ho 1,2,3
# Springer Science+Business Media New York 2015
Abstract Background It is uncertain whether bariatric surgery can be safely performed in secondary hospitals without on-site intensive care unit (ICU) support. This study describes the outcomes of elective bariatric surgery patients who required inter-hospital transfers for unplanned ICU management, extrapolating this as a parameter for secondary hospital safety after bariatric surgery. Methods This was a retrospective, statewide, populationbased, linked data cohort study capturing all adult bariatric surgery patients for an entire Australian state between 2007 and 2011 (n=12,062) with minimum 12-month follow-up. Results In secondary hospitals, 2663 (22.1 %) bariatric patients were operated on, with the majority (n=2553) undergoing sleeve gastrectomies (SG) or adjustable gastric bands (LAGB). Forty-two patients (including 19 LAGB and 20 SG) required inter-hospital transfer to a tertiary hospital for unplanned ICU care (1.6 %, 95 % confidence interval 1.2– 2.1), mainly due to surgical complications. Inter-hospital transfers incurred two deaths, both following sleeve gastrectomies. When compared to patients requiring unplanned ICU admissions after bariatric surgery in tertiary hospitals with an on-site ICU (n = 155), there was no difference in their
* David J. R. Morgan [email protected]
Department of Intensive Care Medicine, St John of God Hospital Subiaco, 12 Salvado Road, Subiaco, Western Australia 6008, Australia
School of Population Health, University of Western Australia, Nedlands, Western Australia, Australia
School of Veterinary and Life Sciences, Murdoch University, Murdoch, Western Australia, Australia
demographic parameters, comorbid illnesses, or mortality (4.8 vs 3.9 %, p=0.68). The mortality following bariatric procedures both statewide (0.2 %) and in secondary hospitals (0.2 %) was both uncommon and comparable. Conclusions Statewide inter-hospital transfers for unplanned ICU care from secondary hospitals were low. Inter-hospital transfer mortality was comparable to a similar bariatric cohort requiring unplanned ICU care after surgery in a tertiary hospital. This suggests that certain bariatric procedures can be safely done in most secondary hospitals where elective ICU admission is deemed unnecessary. Keywords Obesity . Critical care . Hospital safety . Secondary hospital . Outcomes
Introduction Obesity is a chronic, societal health burden leading to an increased prevalence of insidious medical comorbidities and a reduction in national economical productivity [1–3]. A multidisciplinary bariatric surgical approach is currently the most effective treatment for the burgeoning problem of obesity with emerging short- and long-term data supporting both its efficacy and safety [4, 5]. In response to increasing obesity rates, there has been a sustained rise in the incidence of bariatric surgery, with improving surgical and anesthetic techniques inevitably leading to an expanding number of smaller and regional medical facilities offering weight reduction surgery to an increasingly complex patient cohort. With bariatric surgical patients being considered at higher perioperative risk than comparative non-obese patients, critical care support is considered to be an important component in the advanced planning of a bariatric patient’s surgical journey. The increased number and complexity of bariatric patients
inescapably mean there will be a growing minority of patients who will suffer from substantial premorbid comorbidities, or develop significant complications arising from their bariatric surgery, necessitating either elective or emergent intensive care (ICU) support. With reported ICU admission rates of 4– 21 % [6–9], it is questionable whether weight reduction surgery should be offered at a medical facility without the immediate backup of an appropriately staffed on-site ICU. Using this rationale, we set out to determine whether performing bariatric surgery in Western Australian hospitals without onsite ICU support was safe and advisable against the additional risk, inconvenience, and expense of performing an interhospital transfer in an acutely ill patient.
as well as the Western Australian Department of Health Human Research Ethics Committee. For this type of study, formal consent is not required. Definitions
This was a retrospective, multicenter, observational cohort study including all patients who underwent bariatric surgery, with particular scrutiny conferred to high-risk bariatric patients admitted to any intensivist-run ICU and their mode of ICU referral, in the state of Western Australia over a 5-year period between January 1, 2007, and December 31, 2011. All patients were observed until December 31, 2012, unless they died beforehand, allowing for a minimum 12-month followup period after bariatric surgery. In 2011, Western Australia had a population of 2.39 million, comprising 10.4 % of the total Australian population .
Inter-hospital transfer for ICU care was chosen as the primary parameter of bariatric safety in this study due to the resourceconsuming, time-sensitive nature and marker of severity of illness this outcome represents, with existing evidence that acute inter-hospital transfers can result in worsening outcomes [14, 15]. Inter-hospital transfers were defined as patient relocation from one hospital to another hospital for the express purposes of the ongoing management of an unresolved bariatric issue or as a result of an accompanying anesthetic or medical decline during the postoperative period. For the purposes of this study, secondary hospitals are defined as hospitals without an intensivist-operated ICU, while tertiary hospitals were defined as one of the eight Western Australian hospitals with an intensivist-run ICU. Rural/regional hospitals were defined as those being greater than 160 km (~100 mi) from the capital city Perth and thus necessitating an aeromedical evacuation. Patient comorbidities were defined through the documentation of an actively managed condition by a treating physician in the medical records. Severity of ICU illness was defined by the admission Applied Physiological and Chronic Health Evaluation (APACHE) II score which has been previously demonstrated to maintain a similar discrimination as the worst 24-h APACHE II score [16, 17].
Data Sources and Participants
Variables and Outcomes
Participants were identified through the manual screening of eight specialist-run ICU admission databases. All patients older than 18 years were included in the study cohort if they were admitted to an ICU after bariatric surgery or as a result of a subsequent complication directly arising from bariatric surgery. Upon the identification of all appropriate ICU patients, their hospital medical records were reviewed by one of the authors using a preformatted data extraction tool. The above ICU patient database was then merged with an administrational database of all bariatric admissions for the same jurisdiction and time period created by the Western Australian Department of Health Data Linkage Unit (DLU). The DLU systematically connects and updates all the available health data from eight core data elements for every individual within the entire state of Western Australia using the tenth edition international classification of diseases (ICD-10) diagnostic and procedure codes [11, 12]. The accuracy of this database has been previously validated across different jurisdictions applied to other Australian states . Prior ethics approvals were granted from all eight participating hospitals
Primary outcomes included the incidence and reason for bariatric surgery patients requiring inter-hospital transfer where ICU admission and support was a necessary component of the ongoing care. Secondary outcomes included mortality and a comparison against unplanned ICU admissions to tertiary hospitals in an attempt to compare preoperative, perioperative, and postoperative factors for significant unforeseen deviations in patient care between tertiary hospitals and those secondary hospitals without specialist ICU support.
Materials and Methods Study Design and Setting
Statistical Analysis Categorical variables were described in absolute numbers and percentages with comparisons performed using the chisquared test. Continuous variables were described in mean, standard deviation, median, and interquartile range with comparisons performed using the unpaired Student t test. A p value of less than 0.05 was considered statistically significant. Missing data was rare (