To the Editor
Within the VA system, ambulatory surgery centers (ASCs) are gradually replacing traditional hospitals for outpatient surgical care.1 The expansion of ASCs is expected to continue as outpatient surgery increases. Although private sector studies show that patient outcomes are better in ASCs than hospital outpatient departments (HOPDs),2–4 these settings have different financial incentives and a generally healthier population than the VA healthcare system. We examined postoperative outcomes (i.e., specifically hospital admissions) after inguinal hernia surgery in VA ASCs versus HOPDs.
Methods
We used the VA Corporate Data Warehouse (CDW) FY2012-2014 outpatient procedure file to identify surgeries with a CPT code ‘49505’ for repair of initial inguinal hernia. We obtained patient demographics, comorbidities, distance to VA hospital, median income and hospital bed supply in area of residence, ASA score, facility’s geographic region and month/year of the hernia surgery from the CDW and the Area Health Resource File as model predictors. Facilities were excluded if they performed <30 hernia surgeries in FY12-14; cases were excluded if the patient lived >120 miles from a VA hospital or the surgery was emergent. Our final sample included 109 HOPDs and 18 ASCs. The VA Boston Healthcare System’s Institutional Review Board approved this study and waived informed consent.
Postoperative hospitalization was defined as an admission 1-14 days after outpatient surgery with length of stay >24hrs or same day admission with length of stay >48hrs (to exclude observation stays). We performed a chi-square test of proportion and logistic regression predicting admission by ASC adjusting for the predictors described above. To control for factors associated with having surgery in an ASC, we used the same set of predictors in a model estimating likelihood of ASC-based hernia surgery and obtained a propensity score for each patient. We matched ASC patients to those with surgery in a HOPD by propensity score and estimated a second logistic regression model using only ASC status to predict postoperative admission.5 Analyses were performed using SAS.
Results
Of 26,019 outpatient hernia surgeries, 6.9% were performed in an ASC and 1.8% had an admission 0-14 days postoperatively. There were significantly fewer admissions among ASC patients than HOPD patients (1.1% versus 1.9%, p=0.02; Table 1). The initial logistic regression model measured no significant relationship between ASC surgery location and probability of admission (Table 2). The propensity score model included 1,709 matched ASC and HOPD cases; ASC surgery location remained an insignificant factor in predicting admission (odds ratio 0.76, 95% CI: 0.42, 1.38).
Table 1:
Frequency of Patient, Provider and Facility Characteristics in 26,019 VA Hernia Surgeries (FY12-14)
| Type of Predictor | Variable | HOPD (n=109, 24,216 surgeries) N (%) or Mean (SD) | ASC (n=14, 1,803 surgeries) N (%) or Mean (SD) |
|---|---|---|---|
| Outcome | Postoperative Admission | 451 (1.9%) | 20 (1.1%) |
| Patient demographics | Age (years) | 63.5 (12.2) | 62.0 (11.9) |
| Female | 108 (0.5%) | 9 (0.5%) | |
| Race=white | 16,311 (67.4%) | 1,300 (72.1%) | |
| Race=nonwhite | 6,607 (27.3%) | 429 (23.8%) | |
| Race=unknown | 1,298 (5.4%) | 74 (4.1%) | |
| Service connected disability | 595 (2.5%) | 10 (0.6%) | |
| Distance (miles) to nearest VA hospital | 33.7 (37.5) | 44.7 (30.1) | |
| FY12 area median per capita income (USD) | 41,156.4 (9,954) | 40,069.0 (8,914.9) | |
| No zip code in VA patient data | 2,080 (8%) | 91 (0.4%) | |
| Patient comorbidities | No comorbidities | 3,897 (16.1%) | 295 (16.4%) |
| Patient # of comorbidities=1 | 5,384 (22.2%) | 440 (24.4%) | |
| Patient # of comorbidities =2 | 5,058 (20.9%) | 399 (22.1%) | |
| Patient # of comorbidities =3 | 3,768 (15.6%) | 271 (15%) | |
| Patient # of comorbidities =4 | 2,513 (10.4%) | 182 (10.1%) | |
| Patient # of comorbidities =5 | 1,576 (6.5%) | 107 (5.9%) | |
| Patient # of comorbidities =6 | 909 (3.8%) | 49 (2.7%) | |
| Patient # of comorbidities ≥7 | 1,111 (4.6%) | 60 (3.3%) | |
| AIDS | 17 (0.1%) | 255 (14.1%) | |
| Alcohol abuse | 2,725 (11.3%) | 8 (0.4%) | |
| Alcohol liver disease | 176 (0.7%) | 1 (0.1%) | |
| Blood loss anemia | 60 (0.3%) | 222 (12.3%) | |
| Cardiac arrhythmia | 4,066 (16.8%) | 25 (1.4%) | |
| Congestive heart failure | 890 (3.7%) | 279 (15.5%) | |
| Chronic pulmonary disease | 4,345 (17.9%) | 27 (1.5%) | |
| Coagulopathy | 451 (1.9%) | 121 (6.7%) | |
| Deficiency anemia | 2,074 (8.6%) | 323 (17.9%) | |
| Depression | 3,819 (15.8%) | 39 (2.2%) | |
| Diabetes -complicated | 767 (3.2%) | 224 (12.4%) | |
| Diabetes -uncomplicated | 3,662 (15.1%) | 149 (8.3%) | |
| Drug abuse | 1,845 (7.6%) | 70 (3.9%) | |
| Fluid and electrolyte disorders | 1,176 (4.9%) | 968 (53.7%) | |
| Hypertension | 13,327 (55%) | 86 (4.8%) | |
| Hypothyroidism | 1,419 (5.9%) | 22 (1.2%) | |
| Liver disease | 518 (2.1%) | 4 (0.2%) | |
| Lymphoma | 191 (0.8%) | 5 (0.3%) | |
| Metastatic cancer | 180 (0.7%) | 183 (10.2%) | |
| Obesity | 1,949 (8.1%) | 48 (2.7%) | |
| Other neurological disease | 718 (3%) | 45 (2.5%) | |
| Renal failure | 824 (3.4%) | 8 (0.4%) | |
| Paralysis | 131 (0.5%) | 27 (1.5%) | |
| Peptic ulcer bleed | 366 (1.5%) | 205 (11.4%) | |
| Psychoses | 2,602 (10.7%) | 243 (13.5%) | |
| PTSD | 2,825 (11.7%) | 2 (0.1%) | |
| Pulmonary circulation disorders | 158 (0.7%) | 28 (1.6%) | |
| Rheumatoid arthritis | 402 (1.7%) | 43 (2.4%) | |
| Valvular disease | 856 (3.5%) | 57 (3.2%) | |
| Weight loss | 922 (3.8%) | 23 (1.3%) | |
| Procedure characteristics | ASA Score | 2.5 (0.7) | 2.4 (0.6) |
| No ASA (not in NSO surgical data)* | 513 (2.1%) | 841 (46.6%) | |
| Facility characteristics | Facility outpatient surgery volume | 6,523.2 (3,027.1) | 3,185.9 (1,534.6) |
| FY12 hospital beds in area | 2,606.4 (4,554.1) | 1,126.9 (1,662.1) | |
| Atlantic | 5,539 (22.9%) | 315 (17.5%) | |
| Southeast | 4,677 (19.3%) | 343 (19%) | |
| Midwest | 4,641 (19.2%) | 623 (34.6%) | |
| Continental | 5,366 (22.2%) | 204 (11.3%) | |
| Pacific | 3,993 (16.5%) | 318 (17.6%) |
Note, temporal differences averaged 2.7% for HOPDs and ASCs over the course of the study; variation in rates of surgery per month/year was greater for ASCs than HOPDs. These data are available on request.
ASA scores are only available in National Surgery Office (NSO) collected data.
Table 2:
Logistic Regression Results Predicting Postoperative Admissions in 26,019 VA Hernia Surgeries (FY12-14, 109 VA Hospitals, 14 VA ambulatory surgery centers (ASCs))
| Significant Predictors | β Coefficent (Standard Error) | Odds Ratio (95% Confidence Interval) |
|---|---|---|
| ASC | −0.11 (0.25) | 0.9 (0.56-1.45) |
| Age 1st quintile | REF | REF |
| 5th quintile | 0.82 (0.19) | 2.27 (1.55-3.33) |
| Distance (miles) to nearest VA Medical Center 1st quintile | REF | REF |
| 5th quintile | −0.4 (0.19) | 0.67 (0.46-0.97) |
| FY12 area median per capita income (US Dollars) 1st quintile | REF | REF |
| 3rd quintile | −0.41 (0.18) | 0.66 (0.47-0.94) |
| No comorbidities | REF | REF |
| Patient # of comorbidities =4 | 0.71 (0.36) | 2.04 (1.01-4.14) |
| Patient # of comorbidities =6 | 0.95 (0.47) | 2.6 (1.04-6.48) |
| Cardiac arrhythmia | 0.43 (0.13) | 1.53 (1.19-1.97) |
| Duration of Surgery 1st quintile | REF | REF |
| 5th quintile | 0.9 (0.16) | 2.46 (1.79-3.39) |
| FY12 hospital beds in area 1st quintile | REF | REF |
| 2nd quintile | −0.35 (0.17) | 0.71 (0.51-0.98) |
| Surgery performed in Oct FY12 | REF | REF |
| Surgery performed in Feb FY14 | −1.35 (0.59) | 0.26 (0.08-0.82) |
Note: Bolded value indicates significant variable in model. Model performance: pseudo r2=0.09, c-statistic=0.76. Predictors in the logistic regression model that were not significantly associated with postoperative admission include: sex, race, age quintiles 2-4, service connected disability, distance from VAMC quintiles 2-4, FY12 area median per capita income 2nd, 4th and 5th quintile, patient with 1-3, 5, 7 total comorbidities, individual comorbidities (e.g., AIDS) excepting cardiac arrhythmia, duration of surgery quintiles 2-4, ASA score, facility surgical volume, FY12 area hospital bed supply quintiles 3-5, facility’s geographic region, and month/years of the hernia surgery Nov FY12 – Sep FY14 excepting Feb FY14.
Discussion
Private sector studies suggest ASCs are safer than hospitals for routine outpatient surgery, although these findings have been criticized for failing to consider a patient’s likelihood of using an ASC. After controlling for patient characteristics and facility factors in a propensity score matched model, we found no evidence that risk of postoperative admission after hernia surgery differed between ASCs and HOPDs. Although this is a promising sign that VA ASCs represent a safe alternative setting for outpatient surgery, more research is needed to confirm our findings since we did not incorporate non-VA hospitalizations as outcomes. We may also be missing important predictors of having care in an ASC (e.g., patient preference) and experiencing a postoperative admission (e.g., provider availability). Finally, despite its use in private sector quality comparisons, postoperative admissions may be an inadequate marker of patient safety and quality differences between ASCs and HOPDs. Our results show that postoperative admissions after inguinal hernia surgery are rare events regardless of the location of the surgery. Future research should add other types of outpatient surgeries and examine additional outcomes such as postoperative emergency room visits and cost.
Acknowledgment
Dr. Mull had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. We thank Dr. Qi Chen, MD, PhD, VA Boston, for initial methodological input. This research was supported by funding from the VA Health Services Research and Development Service grant number CDA 13-270 (P.I. Mull). The funder was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The authors report no conflict of interests.
Footnotes
Presented at the 2016 Association of VA Surgeons Annual Meeting; abstract #34
References
- 1.Department of Veterans Affairs. VHA DIRECTIVE 2011-037, Facility Infrastructure Requirements to Perform Invasive Procedures in an Ambulatory Surgery Center. Available at: https://www1.va.gov/vhapublications/ViewPublication.asp?pub_ID=2452.2011.
- 2.Chukmaitov AS, Menachemi N, Brown LS, Saunders C, Brooks RG. A comparative study of quality outcomes in freestanding ambulatory surgery centers and hospital-based outpatient departments: 1997-2004. Health Serv Res. 2008;43(5 Pt 1):1485–1504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Paquette IM, Smink D, Finlayson SR. Outpatient cholecystectomy at hospitals versus freestanding ambulatory surgical centers. J Am Coll Surg. 2008;206(2):301–305. [DOI] [PubMed] [Google Scholar]
- 4.Crawford DC, Li CS, Sprague S, Bhandari M. Clinical and Cost Implications of Inpatient Versus Outpatient Orthopedic Surgeries: A Systematic Review of the Published Literature. Orthop Rev (Pavia). 2015;7(4):6177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pizer SD. An intuitive review of methods for observational studies of comparative effectiveness. Health Services and Outcomes Research Methodology. 2009;9(1):54–68. [Google Scholar]