Abstract
Aims: To examine the impact of perioperative inhaled corticosteroids (ICS) on length-of-stay (LOS) and costs among patients receiving high-respiratory-risk surgeries.
Methods: Adult patients who underwent high-respiratory-risk surgeries in 2015 were identified in the Tianjin Urban Employee Basic Medical Insurance database. Patients were grouped into ICS or non-ICS cohorts according to whether they received ICS during the perioperative period of the index hospitalization. Propensity Score Matching was performed to create matched pairs between two cohorts. The impact of perioperative ICS on LOS and direct medical costs was estimated by negative binomial model and generalized liner model.
Results: Eight hundred and twenty-one hospital stays with high-respiratory-risk were selected in the ICS cohort and another 821 stays in the non-ICS cohort were matched. The mean LOS was 13.0 (±0.3) days in the ICS cohort, which was significantly lower than the matched non-ICS cohort. Patients with thorax and ear-nose-throat surgeries had a significant decrease in LOS in the ICS cohort compared to the non-ICS cohort, with a mean decrease of 5.5 and 1.1 days, respectively. In adjusted analyses, perioperative ICS treatment was associated with shorter LOS, lower total, and respiratory-related costs (reductions of 10.1%, 7%, and 5.3%, respectively) after controlling for demographic, clinical, and hospital characteristics.
Limitations: Some respiratory risk factors such as living behavior and environment were unable to be captured and respiratory-related costs might be underestimated, limited by claim data. Lastly, caution should be taken when generalizing the results to other populations, as only patients with moderate-to-severe surgeries on the thorax and above were selected in this study.
Conclusions: Perioperative ICS treatment was associated with decreased LOS and lower costs for patients undergoing high-respiratory-risk surgeries in China.
Introduction
Postoperative respiratory adverse events are one of the most common complications in the surgeries with the incidence ranging from 2–40%Citation1,Citation2. Common respiratory complications include pneumonia, aspiration pneumonitis, respiratory failure, reintubation within 48 h, weaning failure, pleural effusion, atelectasis, bronchospasm, and pneumothorax. Mortality associated with these complications varies, and it can be as high as 48% in cardiac surgeriesCitation3,Citation4. The incremental costs due to respiratory complications range from $7,233 to $1,20,579, which is ∼2–12-fold of the total medical costs in patients without those complicationsCitation5–7.
Respiratory risk factors include advanced age, near-airway surgical sites, prolonged surgery, general anesthesia, intraoperative and postoperative hemorrhage, tracheal intubation, mechanical ventilation, chronic obstructive pulmonary disease (COPD), etc.Citation4,Citation8,Citation9. Corticosteroids, antibacterials, mucolytics as well as bronchodilators have been recommended to be used in high respiratory risk surgeries that can prevent postoperative respiratory complications. Budesonide is one of the most widely used inhaled corticosteroids, which could be delivered locally to the lungs and prevent respiratory complications effectively with a low doseCitation10,Citation11.
Several studies have demonstrated the use of inhaled corticosteroids (ICS), especially inhaled budesonide, could potentially improve lung function and reduce respiratory complicationsCitation12,Citation13. However, these studies were only focused on lung surgeries, and the results were not able to be generalized to the general population due to limited sample size in a single center. In China, only one study evaluated the effect of perioperative ICS on throat-related complicationsCitation14. In addition, there was no real-world evidence about the impact of perioperative ICS on hospital length-of-stay (LOS) and costs. Given that hospital length-of-stay and medical costs play more roles on physicians and patients decision-making nowadays, especially in China, which has severe conflict of increasing medical needs but scarce healthcare resources, it is important to understand the economic value of perioperative ICS in the high-respiratory-risk surgeries. In addition, whether the impact of perioperative ICS on lung function and complications would reflect in the way of LOS and costs is also an unexplored issue.
The objective of this study was to investigate the impact of perioperative ICS on the hospital length-of-stay and direct medical costs in surgeries with high respiratory risks in a real-world setting in China.
Materials and methods
Data source
This retrospective database analysis used de-identified health insurance claims data for individuals enrolled in the Tianjin Urban Employee Basic Medical Insurance (UEBMI), approved by Tianjin Medical Security Bureau. UEBMI is one of two basic medical insurances provided for all employees and retirees in all public and private sectors in China. By 2015, there were 5.22 million enrollees in the Tianjin UEBMI database. This database contains medical and pharmacy claims of insured beneficiaries, including dates of service/prescription, diagnoses, procedures, quantity and cost of services/drugs, as well as patient demographics, enrollment history, dates of death, etc. The analytical sample in this study was a random selection of 30% of all enrollees in 2015. The Safety and Ethics Committee of the School of Pharmaceutical Science and Technology in Tianjin University waived the requirement of ethical approval for the present study.
Study population
Patients who underwent high-respiratory-risk surgeries during January 1, 2015 to December 31, 2015 were included in this study. High-respiratory-risk surgeries were defined as surgeries conducted on the brain, ear-nose-throat (ENT), orofacial (except ENT), thyroid, and thoraxCitation4,Citation12. These surgeries were identified based on surgery name (Supplementary Appendix) and inpatient department (e.g. brain, neurosurgery, ENT, eye, endocrinology, breast, cardio and thoracic department) from the dataset by two clinicians independently. Surgeries were then classified into complexity levels (1–4) based on the Surgical Classification List enacted by the Chinese Ministry of Health in 2011Citation12. The list was an official surgical classification guideline, ranking surgeries into four complexity levels, in which a higher grade indicates more complexity. Only Grade 2–4 surgeries were included in this analysis.
The index date was set as the date that patients received that surgery. Patients in the study sample had to be 18 years old or greater on the index date and be continuously enrolled in the health plan for 1 year before the index date. Patients were grouped into either ICS or non-ICS cohort based on whether they received ICS (i.e. inhalative budesonide, Pulmicort respulas) treatment during the perioperative period. The perioperative period was defined as 7-days prior to and 12-days after the index surgeryCitation13. Patients who were prescribed with ICS outside the perioperative period were excluded from the non-ICS cohort to avoid any potential spillover effect. Those who received an unclear type of anesthesia, since anesthesia type is an important risk factor of respiratory complications, or died during the hospitalization, which may add potential noise in the study cohort with very severe complications, were also excluded.
Patient characteristics
Patient demographic characteristics included age and gender. A Charlson Comorbidity Index (CCI) score was computed for each individual using an algorithm provided by Quan et al.Citation14. Scores may range from 0–33, with a higher score associated with higher probability of mortality or resource use.
Surgical invasive degree, anesthesia type, and duration of mechanical ventilation of both ICS and non-ICS patients were reported in this study. Volume of blood transfusion (including plasma and blood components) during 7 days after the surgery was used as a proxy variable for intro-operative hemorrhage volume, given the latter variable was highly related with damage in surgery but unavailable in claim data. Use of short-acting beta-agonist (SABA)/short-acting muscarinic antagonists (SAMA), and ambroxol were also considered in the analysis to reduce the bias confounded by other airway management medications, following the recommendation in the Chinese Expert Consensus Statement on Perioperative Airway ManagementCitation9.
Besides the clinical characteristics, tier (i.e. community health centers, regional hospitals, or district hospitals) and type (i.e. general or specialist) of hospital were also controlled, because ICS use patterns are different among hospitals in China. Effects of these characteristics were also reported to insight future studies.
Outcomes
Hospital length-of-stay and direct medical costs in the index hospitalization were estimated for both ICS and non-ICS cohorts. Total direct costs comprised surgery costs, procedure costs, laboratory and diagnostic costs, medication costs, medical consumables costs, blood transfusion costs, inpatient bed fees, and others (nourishment, patient accompany, air conditioning and heating fees). Respiratory-related treatment costs were also estimated, which included oxygen inhalation, tracheotomy care, assisted ventilation, and other airway tests and treatment, identified from the Tianjin Diagnosis and Treatment Reimbursement List. All the costs were converted to US dollars ($1 = ￥6.2284).
Statistical analysis
Propensity score matching (PSM) was applied to balance the observable differences at baseline between the ICS and non-ICS groups. A propensity score was estimated for each individual by a logistic regression model using age, gender, CCI, surgery grade and site, anesthesia type, surgery invasion, duration of mechanical ventilation, transfusion, other respiratory management drugs, and hospital characteristics. One-to-one nearest matching without replacement (caliper value was 0.005) was performed to match the non-ICS cohort to the ICS cohort. Student’s t-test or Wilcoxon rank-sum test for continuous variables and Chi-square test or Fisher’s exact test for categorical variables were used to compare patient/clinical characteristics between the two cohorts before and after matching.
Descriptive analyses were conducted by reporting mean, standard deviation (SD), as well as median and range of outcomes for continuous variables, while frequency and percentage were used for categorical variables in this study. After matching, the net effect of ICS on hospital LOS and direct medical cost were evaluated using regression models and adjusting for the same set of covariates used for matching to control their impact on the outcomes. The negative binomial model was used for hospital LOS as it was a limited non-negative variable, and the generalized linear model (GLM) with log-link function and gamma distribution was used for medical cost, given the non-negative and positively skewed distribution of costsCitation15.
All data were processed and analyzed using STATA version 13.0. All statistical tests were two-tailed, and a p-value <.05 was considered statistically significant.
Results
Baseline characteristics
A total of 6,109 hospital stays with high-respiratory-risk surgeries were identified from the dataset, where 821 stays were assigned to the ICS cohort and 821 stays were matched from the non-ICS cohort (). There were significant differences between the two cohorts before matching, where patients in the ICS cohort had higher respiratory risks implied by a higher percentage of general anesthesia, more volume of blood transfusion, and longer mechanical ventilation (). After one-to-one PSM, no significant differences were observed in all baseline characteristics between ICS and non-ICS cohorts ().
Figure 1. Sample selection flow chart.
Table 1. Baseline characteristics before and after PSM.
The study sample had a mean age of 54 years (ICS vs non-ICS: 54.1 vs 53.8, p = .66), and 43% were males (43.6% vs 42.8%, p = .73). Over 50% of patients received ENT surgeries (51.9% vs 54.1%), about 20% received thoracic surgeries (22.9% vs 20.8%). Of the rest, 14%, 6%, and 3% of patients received thyroid, brain, and orofacial surgeries, respectively. The majority of patients (80%) in the study sample underwent Grade 3 or 4 surgeries, and received general anesthesia (82%). Patients received mechanical ventilation for 9 h on average and 5% of them were transfused 0–2,000 ml plasma.
More than 70% of patients received perioperative ICS treatment in general hospitals, and the majority (88%) of those are Tier 3 hospitals in both ICS and non-ICS cohorts.
Comparisons of LOS and direct costs
The hospital LOS between the ICS and non-ICS cohorts by surgical site are reported in . The mean (SD) LOS were significantly lower in the ICS cohort (13.0 (±0.3)) compared to that of 14.8 (±0.4) days in the non-ICS cohort (p < .01). Patients who underwent thoracic and ENT surgeries benefited most from perioperative ICS treatment where the average LOS in ICS cohort were 5.5 days and 1.1 days less than that in the non-ICS cohort, respectively (p < .01). LOS was also reduced by 5 and 2.3 days in ICS patients with brain and orofacial surgeries, but not statistically significant, which may be due to the small sample size (49 vs 48, 24 vs 21, respectively).
Table 2. Hospital length-of-stay between ICS and non-ICS cohort by surgical site.
Mean and median costs in two cohorts were described in . The total direct costs were $6,147 and $5,804 in the ICS and non-ICS cohorts, respectively. However, the difference was not statistically significant (p = .08). The inpatient bed fee was statistically significantly lower in the ICS cohort, compared to in the non-ICS cohort. In most of the cost components such as medication, surgery, laboratory and diagnose, procedure, medical consumables, and blood transfusion costs as well as respiratory-related treatment costs were not significantly different between ICS and non-ICS cases.
Table 3. Total direct medical costs between ICS and non-ICS cohorts.
Impact of ICS on LOS and direct medical costs
The result from Negative Binomial Regression of LOS was consistent with descriptive group comparison. Perioperative ICS treatment was shown to be associated with 10.1% decreased inpatient days (p < .01), after controlling for demographic, clinical, and hospital characteristics. Among these control variables, age, surgery site, mechanical ventilation duration, plasma transfusion, respiratory management drugs, and hospital type also showed a significant impact on LOS ().
Table 4. Impact of ICS on hospital LOS.
From the generalized liner model, perioperative ICS treatment decreased total costs by 5.3% and respiratory-related treatment costs by 7.0% (). A total of $353 was saved by ICS treatment, of which respiratory-related treatment costs saved contributed to 14.4% ($53). Besides, increased total and respiratory related costs were also associated with increased age, male, brain and thorax surgeries, general anesthesia use, long time mechanical ventilation, large volume of blood transfusion, ambroxol use, and higher tier of hospital.
Table 5. Impact of ICS on total direct cost and respiratory-related costs.
Discussion
To our best knowledge, this is the first study to investigate the real-world impact of perioperative ICS for patients undergoing high respiratory risk surgeries around the world. It provided the evidence for the economic value of ICS treatment in high-respiratory-risk surgeries. The present study demonstrated that perioperative ICS treatment could not only shorten LOS by 10.1% in surgeries with high respiratory risks, but also be associated with 5.3% and 7.0% decreases in total costs and respiratory-related treatment costs, respectively.
Clinical benefits of perioperative corticosteroids or ICS on reducing respiratory complications in thoracic, especially lung and nasal, surgeries have been reported by several studiesCitation14–17. A meta-analysis containing 18 randomized controlled trials found pre-operative use of local and/or systemic corticosteroids in functional endoscopic sinus surgery resulted in significantly less blood loss, a shorter operative time, and improved surgical field quality, while postoperative corticosteroids improve postoperative endoscopic scores in chronic rhinosinusitis (CRS) and recurrence rates in cases of chronic rhinosinusitis with nasal polyps (CRSwNP)Citation16. Bölükbas et al.Citation17 compared the use of tiotropium/formoterol/budenoside with tiotropium/formoterol in conjunction with smoking cessation and chest physiotherapy during lung surgeries, and the results indicated that adding inhaled budenoside was associated with an improvement of lung function and less pulmonary complications in the postoperative period. In a double-blind trial involving 94 patients that evaluated the effect of inhaled budesonide on lung function and the inflammatory response to one-lung ventilation in lobectomy for lung cancer, budesonide treatment was associated with significant better outcomes relative to salineCitation18. In another observational study comparing patients inhaling budesonide with those using lidocaine and saline immediately after elective bullae resection surgery, Wen et al.Citation19 found budesonide aerosol inhalation effectively reduced the throat complications. Our results added to the existing evidence that demonstrated patients who underwent thoracic surgeries and received ICS treatment had 5.5 less hospital days than those who did not receive ICS.
Our results suggest that perioperative ICS treatment significantly reduced the total direct costs and respiratory-related treatment costs by 5.3% and 7.0%, respectively. Reduced respiratory-related costs accounted for 14.4% of the total saved costs. Given that it is arduous to specify some of the common services that assist respiratory treatments in the dataset, only clearly specific airway tests and treatments were included when estimating respiratory-related costs, therefore it is possible that our result underestimated the respiratory-related costs.
In cost analyses, the higher unadjusted mean total costs in the ICS cohort was mainly driven by the medical consumable costs ($2,217 vs $1,747). This might be caused by the higher percentage of thoracic surgeries in the ICS cohort (22.9% vs 20.8%), though the difference was insignificant. The medical consumables (e.g. coronary stent) could be extremely expensive in thoracic surgeries, especially in heart surgeries. After adjusting for the surgical sites, ICS showed a negative impact on total costs as well as respiratory costs.
PSM was able to successfully eliminate baseline differences between the ICS cohort and non-ICS cohort for observed characteristics. Predictors of postoperative pulmonary complications, such as age, thoracic surgery, neurosurgery, head and neck surgery, vascular surgery, use of general anesthesia and perioperative blood transfusion were selected based on recommendations from SmetanaCitation4. After PSM, they were all well balanced between two cohorts. Other matching variables, including surgery complexity, surgery invasive degree, duration of mechanical ventilation, were chosen based on clinical experts’ opinion, as they reflected the level of respiratory injuries during operation. Besides, baseline CCI and use of other respiratory drugs were used as the indictors of potential resource use and respiratory disease severity in the PSM. The results of the present study were credible with the appropriate PSM method.
This study also suffers from several limitations. First, due to the limitations of claims data, we were unable to capture not only the clinical indicators which could directly reflect the effect of ICS on lung function and amended complications, but also some other respiratory risk factors such as patient living behavior and environment, including smoking, obesity, physical activity, and pollution, which may potentially impact the healthcare resource use and total cost. Second, the accuracy of surgical site recognition relied heavily on whether the names of surgery and hospital department were uploaded strictly to the claims dataset. And it was carefully reviewed by checking the consistency of the surgery name and department for every patient. Third, respiratory-related costs might be underestimated by conservative recognition of respiratory-related services. Lastly, only patients with moderate-to-severe surgeries on the thorax and above were selected in this study, and the study sample is based on the Tianjin population. Caution should be taken when generalizing the results to other populations.
Conclusion
This study provides real-world evidence on the use of perioperative ICS in high-respiratory-risk surgeries by exploring the impact of ICS treatment on LOS and direct costs using Tianjin Urban Employee Basic Medical Insurance data. Our study finds that perioperative ICS treatment was associated with decreased inpatient days and lower direct costs for patients who underwent high respiratory risk surgeries in China. The benefit of using ICS in the perioperative period of surgeries to prevent respiratory complications were supported, especially among patients who underwent ENT and thoracic surgeries. Promotion of perioperative ICS treatment is encouraged for high-respiratory-risk surgeries in clinical practices.
Transparency
Declaration of funding
This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of financial/other relationships
The authors report no competing interests. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.
191010_ICS_appendix.docx
Download MS Word (29 KB)Acknowledgements
None reported.
References
- Fleischmann KE, Goldman L, Young B, et al. Association between cardiac and noncardiac complications in patients undergoing noncardiac surgery: outcomes and effects on length of stay. Am J Med. 2003;115(7):515–520.
- Belov IV, Komarov RN, Stepanenko AB, et al. Pulmonary and pleural complications of the descending aorta surgery. Khirurgiia (Mosk). 2013;11:16–19.
- Rahmanian PB, Kröner A, Langebartels G, et al. Impact of major non-cardiac complications on outcome following cardiac surgery procedures: logistic regression analysis in a very recent patient cohort. Interact Cardiovasc Thorac Surg. 2013;17(2):319–326.
- Smetana GW. Postoperative pulmonary complications: an update on risk assessment and reduction. Cleve Clin J Med. 2009;76(Suppl_4):S60–S65.
- Sabate S, Mazo V, Canet J. Predicting postoperative pulmonary complications: implications for outcomes and costs. Curr Opin Anaesthesiol. 2014;27(2):201–209.
- Whitmore RG, Stephen J, Stein SC, et al. Patient comorbidities and complications after spinal surgery: a societal-based cost analysis. Spine. 2012;37(12):1065–1071.
- Cassidy MR, Rosenkranz P, McCabe K, et al. I COUGH: reducing postoperative pulmonary complications with a multidisciplinary patient care program. JAMA Surg. 2013;148(8):740–745.
- Sharafkhaneh A, Falk JA, Minai OA, et al. Overview of the perioperative management of lung volume reduction surgery patients. Proc Am Thorac Soc. 2008;5(4):438–441.
- Zhi XY, He JH, Liu LX, et al. Chinese expent consensus statement on multi-disciplinary perioperative airway management (2016 version). Chin J Clin Thorac Cardiovasc Surg (Chinese). 2016;23:641–645.
- Maltais F, Ostinelli J, Bourbeau J, et al. Effect of inhaled formoterol and budesonide on exacerbations of asthma. New England Journal of Medicine. 1997;337(20):1405.
- Pauwels RA, Löfdahl C-G, Postma DS, et al. Effect of inhaled formoterol and budesonide on exacerbations of asthma. N Engl J Med. 1997;337(20):1405.
- Chinese Ministry of Health, Surgical Classification List (2011). 2011. Available from https://wenku.baidu.com/view/27903c96852458fb760b5619.html
- Zhao MY. Routine of perioperative nursing (Chinese). Beijing: Science Press; 2010.
- Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–1139.
- Chang H-Y, Kharrazi H, Bodycombe D, et al. Healthcare costs and utilization associated with high-risk prescription opioid use: a retrospective cohort study. BMC Med. 2018;16(1):69.
- Pundir V, Pundir J, Lancaster G, et al. Role of corticosteroids in functional endoscopic sinus surgery – a systematic review and meta-analysis. Rhinology. 2016;54(1):3–19.
- Bolükbas S, Eberlein M, Eckhoff J, et al. Short-term effects of inhalative tiotropium/formoterol/budenoside versus tiotropium/formoterol in patients with newly diagnosed chronic obstructive pulmonary disease requiring surgery for lung cancer: a prospective randomized trial. Eur J Cardiothorac Surg. 2011;39(6):995–1000.
- Ju NY, Gao H, Huang W, et al. Therapeutic effect of inhaled budesonide (Pulmicort(R) Turbuhaler) on the inflammatory response to one-lung ventilation. Anaesthesia. 2014;69(1):14–23.
- Wen KL, Yang MC, Jiang R, et al. Budesonide combining Lidocaine aerosol inhalation reduces the double-lumen endotracheal intubation-related throat complications. J Kunming Med Uni (Chinese). 2015;36:105–109.