Skip Navigation LinksHome > December 2013 - Volume 119 - Issue 6 > Patient Selection for Day Case-eligible Surgery: Identifying...
doi: 10.1097/ALN.0000000000000005
Perioperative Medicine

Patient Selection for Day Case-eligible Surgery: Identifying Those at High Risk for Major Complications

Mathis, Michael R. M.D.*; Naughton, Norah N. M.D., M.B.A.; Shanks, Amy M. M.S.; Freundlich, Robert E. M.D., M.S.*; Pannucci, Christopher J. M.D., M.S.§; Chu, YiJia M.D.*; Haus, Jason M.D.*; Morris, Michelle M.S.; Kheterpal, Sachin M.D., M.B.A.

Free Access
University of Michigan
Article Outline
Collapse Box

Author Information

Collapse Box


Background: Due to economic pressures and improvements in perioperative care, outpatient surgical procedures have become commonplace. However, risk factors for outpatient surgical morbidity and mortality remain unclear. There are no multicenter clinical data guiding patient selection for outpatient surgery. The authors hypothesize that specific risk factors increase the likelihood of day case-eligible surgical morbidity or mortality.
Methods: The authors analyzed adults undergoing common day case-eligible surgical procedures by using the American College of Surgeons’ National Surgical Quality Improvement Program database from 2005 to 2010. Common day case-eligible surgical procedures were identified as the most common outpatient surgical Current Procedural Terminology codes provided by Blue Cross Blue Shield of Michigan and Medicare publications. Study variables included anthropometric data and relevant medical comorbidities. The primary outcome was morbidity or mortality within 72 h. Intraoperative complications included adverse cardiovascular events; postoperative complications included surgical, anesthetic, and medical adverse events.
Results: Of 244,397 surgeries studied, 232 (0.1%) experienced early perioperative morbidity or mortality. Seven independent risk factors were identified while controlling for surgical complexity: overweight body mass index, obese body mass index, chronic obstructive pulmonary disease, history of transient ischemic attack/stroke, hypertension, previous cardiac surgical intervention, and prolonged operative time.
Conclusions: The demonstrated low rate of perioperative morbidity and mortality confirms the safety of current day case-eligible surgeries. The authors obtained the first prospectively collected data identifying risk factors for morbidity and mortality with day case-eligible surgery. The results of the study provide new data to advance patient-selection processes for outpatient surgery.
Back to Top | Article Outline

What We Already Know about This Topic

* Risk factors for major morbidity and mortality from outpatient surgery are not clearly defined
Back to Top | Article Outline

What This Article Tells Us That Is New

* In a review of approximately 250,000 cases in the National Surgery Quality Improvement Program, early perioperative morbidity and mortality occurred in approximately 1:1,000 patients
* Predictors for morbidity and mortality were overweight or obesity, chronic obstructive pulmonary disease, history of transient ischemic attack/stroke, hypertension, previous cardiac surgical intervention, and prolonged operative time
OVER the past 3 decades, the proportion of surgeries performed in an outpatient setting has increased dramatically, comprising 16% of all surgeries in 1980 and steadily growing to include over 50% in the 1990s and over 60% by 2007.1,2 This drive toward outpatient surgery can be attributed to several factors. Improvements in anesthetic care, including innovations such as shorter-acting anesthetic agents and improved cardiopulmonary monitoring, have allowed for fewer adverse anesthetic effects. Innovations in minimally invasive surgical techniques have decreased the need for inpatient hospitalization.3 Economic pressures have also influenced increased adoption of outpatient surgery. Estimates have determined charges per visit for surgical procedures to be approximately five times less for outpatient versus inpatient surgery although unadjusted for length-of-stay or procedural complexity.4
Patient selection and these advances in perioperative care have allowed outpatient surgical procedures to be performed at an exceedingly low rate of morbidity or mortality.5–8 However, concern over patient safety remains, as the outpatient surgical population has increased not only in volume but also in age and complexity, necessitating improved preoperative screening.9–12 In addition, the explosion of ambulatory surgery centers (ASCs) has created a need to identify patients suitable for receiving surgical procedures on an ambulatory basis. There are no national, prospectively collected data regarding optimal patient selection for ASC procedures.13,14 Patient selection is largely guided by administrative data, focused on risk of “readmission” or incidence of complications.5–8,15
Although data are available regarding the incidence of morbidity and mortality after outpatient surgery, there are no prospectively collected clinical data to guide the outpatient surgery–selection processes. Risk factors for adverse outcomes have been proposed,16–18 and calls for such a study have been made, as this would provide valuable information to a clinician determining a patient’s eligibility for surgery on an ambulatory basis.12,19,20 We hypothesize that specific patient history and surgical characteristics place patients scheduled for common day case-eligible surgeries at a greater risk of major morbidity and mortality within 72 h after such procedures.
Back to Top | Article Outline

Materials and Methods

Clinical information on all adult patients (aged ≥18 yr) who underwent surgical procedures from 2005 to 2010 was obtained from the Participant Use Data File of the American College of Surgeons’ National Surgical Quality Improvement Program (ACS-NSQIP). Within this dataset, data are deidentified, contain no protected health information, and are made publicly available. As a result, the proposed analysis was performed under an existing institutional review board exemption (University of Michigan, Ann Arbor, Michigan, HUM22030). Patient consent was waived because no protected health information was available in the dataset.
All operative procedures performed with general, spinal, or epidural anesthesia from more than 250 medical centers across the country are eligible for inclusion in this dataset. Procedures at each contributing site are recorded in 8-day cycles, with the first 40 operations within each cycle included in the database. To ensure data heterogeneity, an 8-day cycle is used to favor each day of the week equally when beginning data collection on subsequent cycles; in addition, high-volume, low-risk cases (e.g., cholecystectomy, inguinal hernia repair) are capped at five cases per cycle. To ensure fidelity of data collection, clinical nurses trained in data collection are assigned at each site, with interrater reliability measurements taken through periodic site reviews. All sites with interrater reliability less than 95% are excluded from the database. Currently, fidelity of data collection has been demonstrated to be excellent, with sites showing less than 1.5% variable disagreement during annual audits.21 Patients included in the database underwent a variety of operations, including general, orthopedic, plastic, urologic, obstetric, gynecologic, and vascular surgical procedures. Patients were followed through their operative course until postoperative day 30. Trained nurses then performed a patient chart review at the institution where the surgical procedure was performed and made direct patient contact to further identify complications diagnosed and treated at other institutions. The methodology for prospective data collection within this database, demonstrating high degrees of accuracy and reproducibility, has been described in previous literature.21–23
For this study, all adult patients who were scheduled as outpatient were eligible for inclusion. Outpatient status for each procedure was determined at participating institutions in compliance with federal guidelines. As confirmed among reviewers across several NSQIP-participating sites, a surgical procedure is defined as “outpatient” if the patient arrived to the ASC or hospital-based facility on the same day as the procedure performed and was discharged the same day; this applied to patients planned for both same-day discharge and “23-h observation” intended for overnight stay, but then discharged before overnight stay.# In addition, this excluded same-day admission patients planned for hospital admission for one or more nights after surgery, patients who underwent surgery on a hospital day subsequent to hospital admission, and finally patients admitted from the emergency department undergoing surgery the same day, and possibly discharged the same day (each defined as “inpatient”).
Table 1
Table 1
Image Tools
Fig. 1
Fig. 1
Image Tools
The ACS-NSQIP Participant Use Data File dataset includes data from freestanding ASCs, ASCs attached to acute care hospitals, and acute care hospitals; there are no data elements allowing differentiation of location of care. “Common” day case-eligible surgeries were defined as the 100 most frequently used Current Procedural Terminology (CPT) codes for which outpatient surgery claims were billed to Blue Cross Blue Shield, Michigan Liability, regular business (commercial Preferred Provider Organization) during the 2010 calendar year, in conjunction with 50 commonly performed procedures in ASCs as determined via a national claims review of Medicare beneficiaries during the 2010 calendar year (fig. 1).24 Day case-eligible surgical patients in the NSQIP-Participant Use Data File file were then filtered to only include these commonly performed CPT codes. A summary of the most commonly observed surgical procedures are listed in table 1. To improve validity of the final analytical dataset, we then excluded cases clearly inappropriately coded as outpatient by removing patients who were ventilator dependent, American Society of Anesthesiologists physical status 5 or 6, experiencing preoperative sepsis, or did not have surgery on the recorded admission date. In addition, due to a low incidence, we excluded patients who had received preoperative transfusions within 72 h of operation or had a tumor involving the central nervous system (fig. 1). A summary of all ACS-NSQIP variables used as inclusion/exclusion criteria are detailed in appendix 1.
Table 2
Table 2
Image Tools
Within the study population, perioperative patient data collected included anthropometric data and the presence of a number of chronic diseases detailed in table 2; specific definitions for these perioperative variables are detailed in appendix 2 and reflect the standard ACS-NSQIP definitions. Among the risk factors selected for the study, many have been previously considered as risk factors for adverse outcomes related to ambulatory surgery, including age,11,15,25,26 sex,25,27 body mass index (BMI),11,28 type of surgery,11,25,27 surgical duration,11,15,25,26,29 and presence of comorbidities including diabetes mellitus,11 chronic obstructive pulmonary disease (COPD),11 congestive heart failure,11,25 previous myocardial infarction,11 previous cardiac surgery,17 hypertension,11 angina,11 peripheral vascular disease,15 renal failure,11 cancer,15 history of stroke,11,15 history of smoking,11,17 and recent surgical admission.10,16 The ACS-NSQIP dataset also included other salient risk factors, such as recent significant weight loss, steroid use, heavy alcohol intake, paraplegia/quadriplegia, and pregnancy. Age was converted from a continuous variable to categorical by establishing age ranges by decade with 18–30 yr considered the reference group. BMI was calculated using the height and weight as reported; only values within a valid range of 10–80 were used for analysis. After BMI calculation, patients were categorized based on the World Health Organization classifications for underweight, normal weight (reference group), overweight, and obese.30 In addition, distinct ACS-NSQIP variables addressing a single clinical concept were collapsed for the purposes of this study. Diabetes mellitus requiring insulin treatment and requiring oral treatment were combined into diabetes. Acute renal failure and current dialysis were combined into renal failure. History of transient ischemic attacks (TIA) and history of cerebrovascular accident (CVA) were combined into history of TIA/CVA. Cancer was defined as a diagnosis of disseminated cancer, radiotherapy within 90 days, and/or chemotherapy within 30 days of operation. Previous percutaneous coronary intervention (PCI) and history of cardiac surgery were combined into previous cardiac surgical intervention.
We defined prolonged operative time as any case duration greater than the CPT code-specific 75th percentile for each surgical case included in the study, as previously defined by the Centers for Disease Control methodology.31,32 To risk-adjust based on the inherent risk of each type of surgery, we computed a surgical complexity score based on a previously established technique using primary CPT code.33 This technique results in one continuous score based on a logistic regression model. As applied to our study, each CPT code was placed into the model with our primary outcome as the dependent variable. Resulting adjusted odds ratios (AORs) were calculated for each CPT code, converted to logarithmic scale, and applied to each surgical case included in our study. The surgical complexity score was then included as a continuous variable for all statistical models used.
Table 3
Table 3
Image Tools
The primary outcome studied was composite perioperative morbidity/mortality, defined as the occurrence of any intraoperative or postoperative major complication occurring within 72 h of surgery, as detailed in table 3. Specific definitions for variables comprising the primary outcome are detailed in appendix 3. Consistent with previous outpatient surgery literature,5 a 72-h outcome window was selected to identify patients experiencing morbidities who may have potentially benefitted from postoperative inpatient monitoring and treatment before any scheduled outpatient postoperative follow-up appointment. Timing of major complications included in the primary outcome definition was defined as the time of definitive diagnosis.
In addition, a secondary outcome studied was unplanned admission, defined as a hospital length-of-stay greater than 1 day—despite being scheduled as outpatient, as prerequisite for inclusion in the patient population studied. This secondary outcome was chosen on the basis that previous literature has focused on this outcome to guide outpatient surgical selection.15,34
To ensure validity of the cases in the final analytical dataset, each of the cases meeting the primary or secondary outcome was manually reviewed by two of the study investigators (M.R.M., M.M., or S.K.) to confirm that no preoperative or postoperative attributes suggested misclassification of the case as outpatient. Specifically, the primary procedure code, additional procedure codes, preoperative comorbidity data, and duration of procedure were reviewed.
Back to Top | Article Outline
Statistical Analysis
Statistical analysis of data was performed using RStudio and SPSS version 21.0 (IBM, Somers, NY). Basic descriptive statistics were calculated for demographic and anthropometric data. Pearson chi-square or Fisher exact tests (for categorical variables) were used to assess baseline univariate clinical differences between patients who did and did not demonstrate the primary outcomes of interest.
Before developing a prediction model to determine independent risk factors, all variables in table 2 were tested for collinearity by investigating the condition index. If the condition index was greater than 30, then pairwise Pearson correlation matrices were constructed to determine which variables were highly correlated. If the condition index was less than 30, then no collinearity was detected, and all variables were eligible for regression model entry. Variables from table 2 which demonstrated a univariate P value of less than 0.20 or clinical relevance were entered into a semi-parsimonious logistic regression model to identify independent predictors of measured outcomes. A multivariate P value of less than 0.05 was considered statistically significant and an independent predictor. Measures of effect size were reported as AORs and 95% CIs. Goodness-of-fit was assessed using the Omnibus Tests of Model Coefficients as well as the Hosmer and Lemeshow Test. Overall predictive capability of each model was assessed using the c-statistic. In addition, the model was validated using bootstrapping, performed in RStudio using the bootcov function with 1,000 replacements. A P value of less than 0.05 was considered statistically significant and an independent predictor in the bootstrapped dataset.
Back to Top | Article Outline


Through a review of the Michigan Blue Cross Blue Shield and the national Medicare 2010 databases, a total of 140 CPT codes were identified as common day case-eligible surgical procedures. Within the NSQIP database, we identified 341,427 cases with one or more matching CPT codes, of which 257,903 were scheduled on an outpatient basis. Among these cases, 244,397 cases met the inclusion criteria for this study as detailed in figure 1. Exactly 241,600 cases (98.9%) involved a length-of-stay less than or equal to 1 day, and data completion rate was noted to be 98% or greater for all but two variables: previous operation within 30 days (94%) and pregnancy (87%).
Within this study population, primary outcome analysis identified 232 cases experiencing an event: a total of 21 mortalities and 234 perioperative morbidities (multiple morbidities in some cases) within 72 h postoperatively. This corresponded to an incidence of 0.095%, or approximately 1 in 1,053 cases. It was noted that no intraoperative deaths occurred; nine occurred on the day of surgery, seven on postoperative day 1, and five on postoperative day 2. Of the 232 cases experiencing an event, 195 (84%) were discharged within 23 h of surgery. Of the 234 perioperative morbidities, the most common events included pneumonia (46), unplanned postoperative intubation (37), wound disruption (25), postoperative bleeding (21), and sepsis (19).
Fig. 2
Fig. 2
Image Tools
A variety of patient comorbidities were identified as significant univariate predictors of 72-h perioperative morbidity or mortality, as described in table 2. On multivariable analysis, collinearity diagnostics did not demonstrate any condition indices greater than 30, and therefore, no variables were removed from the analysis. All variables in table 2 meeting logistic regression model inclusion criteria plus the surgical complexity score were subsequently entered into the semi-parsimonious logistic regression model. Through multivariable analysis of primary outcome, seven independent predictors of perioperative morbidity or mortality were identified while controlling for surgical complexity: overweight BMI (AOR, 1.6; 95% CI, 1.1–2.3), obese BMI (AOR, 2.0; 95% CI, 1.4–3.0), prolonged operative time (AOR, 1.7; 95% CI, 1.3–2.2), previous PCI/cardiac surgery (AOR, 1.7; 95% CI, 1.2–2.6), COPD (AOR, 2.4; 95% CI, 1.4–4.0), hypertension (AOR, 1.7; 95% CI, 1.2–2.3), and history or TIA/CVA (AOR, 2.1; 95% CI, 1.4–3.4). These independent predictors are shown in figure 2. The Omnibus Tests of Model Coefficients demonstrated a chi-square value of 155, 24 degrees of freedom, and a P value less than 0.001. The Hosmer and Lemeshow Test demonstrated a chi-square value of 8.12, 8 degrees of freedom, and a P value of 0.42. The c-statistic was 0.72 (95% CI, 0.69–0.76). All seven of the independent predictors were validated using bootstrapping with 1,000 samples with replacement. The Somers Dxy was 0.4646 for the original data set, 0.4901 for the training set, and 0.4175 for the test set. The optimism was 0.0726 and the bias-corrected Somers Dxy was 0.3920.35
Secondary outcome analysis recognized 2,797 cases (1.1%) described as unplanned admission. An additional multivariable regression model identified 17 independent predictors: male sex (AOR, 0.77; 95% CI, 0.71–0.85), age 51–60 yr (AOR, 1.2; 95% CI, 1.0–1.4), age 61–70 yr (AOR, 1.4; 95% CI, 1.2–1.7), age 71–80 yr (AOR, 1.6; 95% CI, 1.3–2.0), age 81–90 yr (AOR, 2.4; 95% CI, 1.9–3.0), underweight BMI (AOR, 2.0; 95% CI, 1.5–2.7), obese BMI (AOR, 1.4; 95% CI, 1.2–1.6), prolonged operative time (AOR, 3.1; 95% CI, 2.8–3.3), diabetes mellitus (AOR, 1.3; 95% CI, 1.1–1.5), COPD (AOR, 1.4; 95% CI, 1.2–1.8), previous PCI/cardiac surgery (AOR, 1.3; 95% CI, 1.2–1.5), hypertension (AOR, 1.1; 95% CI, 1.0–1.2), renal failure/dialysis (AOR, 2.3; 95% CI, 1.6–3.4), history of TIA/CVA (AOR, 1.5; 95% CI, 1.2–1.8), paraplegia or quadriplegia (AOR, 2.7; 95% CI, 1.5–4.9), current steroid use (AOR, 1.6; 95% CI, 1.2–2.0), and previous operation within 30 days (AOR, 0.58; 95% CI, 0.35–0.94). A chi-square value of 1,158, 30 degrees of freedom, and a P value less than 0.001 were determined by the Omnibus Tests of Model Coefficients. The Hosmer and Lemeshow Test revealed a chi-square value of 14.4, 8 degrees of freedom, and a P value of 0.07. The c-statistic was 0.70 (95% CI, 0.69–0.71).
Back to Top | Article Outline


The expansion of outpatient surgery across the United States is sharply contrasted by a shortage of outcomes data for patients undergoing procedures in this care setting. Current knowledge is limited to case series, single-center data, and administrative data analyses, and this continues to create a demand for evidence-based research to direct future initiatives. To this end, our study sought out specific patient populations for whom the day-case surgical model may pose significant risks. Given the paucity of outpatient data overall—and in particular for ASC patients—these outpatient data serve as a critical starting point for research focused on the ASC care model.
In our study, the incidence of perioperative morbidity or mortality was 0.095% among 244,397 adult outpatients undergoing common day case-eligible surgical procedures, corresponding to 1 in 1,053 cases. We identified seven independent predictors of perioperative morbidity or mortality when controlled for surgical complexity: overweight BMI, obese BMI, prolonged operative time, COPD, hypertension, previous PCI/cardiac surgery, and history of TIA/CVA. Early postoperative pneumonia, unplanned postoperative intubation, and wound disruption were among the most common morbidities identified.
The incidence of perioperative morbidity/mortality identified in our study is consistent with previous single-center literature, which notes major morbidity rates from 0.09 to 0.60% through the perioperative period among common ambulatory surgical procedures.5,7,8 The low rate confirms the relative safety of day case-eligible surgical procedures, aided by current patient-selection processes—although nonuniform—which continue to evolve with emerging literature. Such selection processes have cited case series and single-center analyses suggesting that poorly stabilized cardiopulmonary status, obstructive sleep apnea, advanced age, increased surgical invasiveness, and recent hospitalizations are characteristics placing patients at high risk for major morbidity, death, or hospital admission after outpatient surgery.10,11,16,18 An administrative analysis performed by Fleisher et al.15 did note a 0.58% rate of hospital admission after outpatient surgery characterized by similar risk factors identified in our study although nonspecific for cause of readmission. Our prospectively collected clinical data provide validation to the existing retrospective administrative or single-center literature, with some identified risk factors confirming previous hypotheses, and others perhaps newly elucidated.
In addition to modest discriminating capacity (c-statistic of 0.72), the model demonstrates clinical relevancy because each independent predictor had an AOR of 1.6 or greater for perioperative morbidity within 72 h. Many of the independent predictors had even higher AORs, such as COPD or a history of TIA/CVA. Because the underlying rate of the primary outcome is low, this does not translate to a large absolute risk increase. However, in the current environment of limited clinical data to guide patient selection, these data serve to provide a previously unavailable evidence basis for this process.
Administrative data analyses focused on a readmission outcome have previously noted prolonged operative duration, cerebrovascular disease, obesity, and cardiac disease to increase the risk after outpatient surgery.15,34 Our prospectively collected clinical data confirm similar observations that, until now, were only noted in administrative data. In addition, we have noted hypertension requiring medications and COPD requiring medications as additional risk factors previously not reported in clinical literature. These represent additional preoperative characteristics which a medical director should consider when screening for appropriateness of day case-eligible surgery at an ASC. Interestingly, however, our multivariate regression analysis reported that many variables were not independently associated with an increased risk of perioperative morbidity/mortality: these included sex, advanced age, paraplegia/quadriplegia, cancer, renal failure, steroid use, congestive heart failure, diabetes mellitus, underweight BMI, angina, and heavy alcohol intake.11,15,16 However, due to the low event rate, we may not have been able to observe statistically significant relationships despite an underlying clinically meaningful relationship.
Although our analysis dataset does not specifically identify which cases were performed at an ASC versus acute care hospital, the data do offer incremental insight for an ASC medical director attempting to establish objective criteria for patient selection. Many ASC criteria are based on subjective patient risk assessments such as the American Society of Anesthesiologists physical status classification, which has poor interrater reliability and can be manipulated.36 Although physical status classification has been demonstrated to be an important statistical tool for risk adjustment, its use for clinical decision making has never been established. Our comorbidity-driven analysis provides specific disease definitions usable for a more transparent patient-selection process. The data must be interpreted with caution given the low absolute risk increases involved. Our data do not offer definitive, comprehensive selection criteria for ASC patient selection. Clearly, the low event rate noted in our dataset demands further research using databases specifically designed for analysis of outpatient surgery populations or requires that a specific database of high-risk outpatient surgery patients be created.
Our secondary analysis evaluated the incidence and predictors of unplanned admission after day case-eligible surgery. We observed 2,797 patients (1.1%) required an unplanned admission. Unfortunately, the dataset does not describe the specific event or issue that necessitated an admission, and we do not wish to conjecture using the postoperative events. However, these data are consistent with existing administrative or single-center data evaluating the incidence of admission. The risk model also had modest discriminating capacity with a c-statistic of 0.70 and likely encompasses a wide range of admission reasons. The factors identified are consistent with previous administrative literature.
Our study has several limitations. We sought to achieve an appropriate balance between generalizability and focused clinical results by using the Michigan Blue Cross Blue Shield and Medicare databases to identify common surgeries performed on an outpatient basis. However, given the heterogeneity of outpatient surgery centers throughout the United States, clinicians must use caution in applying these findings. In addition, the central limitation of our study was that although the procedures were scheduled as outpatient, the ACS-NSQIP dataset does not identify whether they were performed at an ASC. Nevertheless, the identified high-risk patient characteristics are still of value to ASC medical directors seeking objective patient-selection criteria. Variations in procedures and data-collection methods may have existed across ACS-NSQIP institutions although it should be noted that this was limited through the use of specific, highly enforced, rigorously validated data-collection protocols. Although attempts to minimize sources of misclassification have included using a database excluding hospital sites with suboptimal interrater reliability, developing exclusion criteria filtering invalid data, and finally a manual review of all primary outcomes included in the dataset, the authors acknowledge that it is possible that inaccuracies in data collection not detectable by any means using all other case data recorded for each primary outcome may have existed. In addition, our study shared the limitations of any observational trial. Within the ACS-NSQIP database, variables are recorded for quality improvement and research purposes, and data beyond this scope are not available. Perioperative risk factors assessed in this study were limited to existing ACS-NSQIP definitions, and those chosen were limited to those hypothesized in previous literature. In addition, ACS-NSQIP definitions do not discriminate severity within a given preoperative comorbidity. The risk factors evaluated were also limited in number as to avoid multivariable regression model overfitting. As demonstrated by c-statistic of 0.72, our multivariable model shows modest predictive capability, with other risk predictors not accounted for. Finally, as data are unavailable as to whether day case-eligible surgical cases included in the study were performed at a hospital-based facility or ASC, overreporting of patient complications may have existed, as it is possible that medical directors within the NSQIP database scheduled procedures as outpatient with the knowledge that outpatient procedures would be performed at a hospital-based facility (with increased access to healthcare resources, and thus higher acceptable patient risk) rather than ASC. Although a consistent method of classifying surgical procedures as “inpatient” versus “outpatient” was verified across multiple institutions, the authors acknowledge that nurses trained in data entry at other participating sites not contacted may have reasonably classified same-day admissions as “outpatient”; for this reason, further overreporting of patient complications may have existed.
Despite these limitations, our study remains the first national clinical analysis assessing risk factors for specific major morbidities and mortality after day case-eligible surgery. In accomplishing this goal, our results can provide ASC medical directors with information beyond just the incidence of these outcomes. In addition, in providing data on specific perioperative morbidities, information more tangible than a risk of inpatient admission can be provided to medical directors. In this sense, our study provides a broader evidence base for patient screening and allows for more specific patient counseling on perioperative risks.
In conclusion, we report major perioperative morbidities or mortalities to occur once in every 1,053 day case-eligible surgical cases in our study population, and they were independently associated with overweight BMI, obese BMI, prolonged operative time, previous PCI/cardiac surgery, hypertension, COPD, and a history of TIA/CVA. In the midst of a rapidly growing, increasingly complex outpatient population, the risk factors identified in our study provide evidence upon which an ASC medical director’s patient-screening decisions can be based.
# Electronic-mail communication with NSQIP data entry nurse specialists and trainers, July 2013. Cited Here...
Back to Top | Article Outline


1. Hospital admissions, average length of stay, outpatient visits, and outpatient surgery by type of ownership and size of hospital: United States, selected years 1975–2007. National Center for Health Statistics. Health, United States, 2009: With Special Feature on Medical Technology, Table 104. 2010

2. Cullen KA, Hall MJ, Golosinskiy A. Ambulatory surgery in the United States, 2006. Natl Health Stat Report. 2009:1–25

3. Fuchs KH. Minimally invasive surgery. Endoscopy. 2002;34:154–9

4. Russo A, Elixhauser A, Steiner C, Wier L Hospital-Based Ambulatory Surgery, 2007: Statistical Brief #86. Rockville, Maryland, Agency for Healthcare Research and Quality, 2006

5. Warner MA, Shields SE, Chute CG. Major morbidity and mortality within 1 month of ambulatory surgery and anesthesia. JAMA. 1993;270:1437–41

6. Hancox JG, Venkat AP, Coldiron B, Feldman SR, Williford PM. The safety of office-based surgery: Review of recent literature from several disciplines. Arch Dermatol. 2004;140:1379–82

7. Engbaek J, Bartholdy J, Hjortsø NC. Return hospital visits and morbidity within 60 days after day surgery: A retrospective study of 18,736 day surgical procedures. Acta Anaesthesiol Scand. 2006;50:911–9

8. Keyes GR, Singer R, Iverson RE, McGuire M, Yates J, Gold A, Thompson D. Analysis of outpatient surgery center safety using an internet-based quality improvement and peer review program. Plast Reconstr Surg. 2004;113:1760–70

9. White PF, White LM, Monk T, Jakobsson J, Raeder J, Mulroy MF, Bertini L, Torri G, Solca M, Pittoni G, Bettelli G. Perioperative care for the older outpatient undergoing ambulatory surgery. Anesth Analg. 2012;114:1190–215

10. Gupta A. Preoperative screening and risk assessment in the ambulatory surgery patient. Curr Opin Anaesthesiol. 2009;22:705–11

11. Bettelli G. High risk patients in day surgery. Minerva Anestesiol. 2009;75:259–68

12. Eichhorn V, Henzler D, Murphy MF. Standardizing care and monitoring for anesthesia or procedural sedation delivered outside the operating room. Curr Opin Anaesthesiol. 2010;23:494–9

13. Pannucci CJ, Shanks A, Moote MJ, Bahl V, Cederna PS, Naughton NN, Wakefield TW, Henke PK, Campbell DA, Kheterpal S. Identifying patients at high risk for venous thromboembolism requiring treatment after outpatient surgery. Ann Surg. 2012;255:1093–9

14. Wynia MK, Classen DC. Improving ambulatory patient safety: Learning from the last decade, moving ahead in the next. JAMA. 2011;306:2504–5

15. Fleisher LA, Pasternak LR, Lyles A. A novel index of elevated risk of inpatient hospital admission immediately following outpatient surgery. Arch Surg. 2007;142:263–8

16. Lermitte J, Chung F. Patient selection in ambulatory surgery. Curr Opin Anaesthesiol. 2005;18:598–602

17. Bryson GL, Chung F, Finegan BA, Friedman Z, Miller DR, van Vlymen J, Cox RG, Crowe MJ, Fuller J, Henderson C. Patient selection in ambulatory anesthesia—An evidence-based review: Part I. Can J Anaesth. 2004;51:768–81

18. Khan M, Ahmed A, Abdullah L, Nizar A, Fareed A, Khan FA. Unanticipated hospital admission after ambulatory surgery. J Pak Med Assoc. 2005;55:251–2

19. Metzner J, Domino KB. Risks of anesthesia or sedation outside the operating room: The role of the anesthesia care provider. Curr Opin Anaesthesiol. 2010;23:523–31

20. Van De Velde M, Kuypers M, Teunkens A, Devroe S. Risk and safety of anesthesia outside the operating room. Minerva Anestesiol. 2009;75:345–8

21. Shiloach M, Frencher SK Jr, Steeger JE, Rowell KS, Bartzokis K, Tomeh MG, Richards KE, Ko CY, Hall BL. Toward robust information: Data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg. 2010;210:6–16

22. Daley J, Khuri SF, Henderson W, Hur K, Gibbs JO, Barbour G, Demakis J, Irvin G III, Stremple JF, Grover F, McDonald G, Passaro E Jr, Fabri PJ, Spencer J, Hammermeister K, Aust JB, Oprian C. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: Results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg. 1997;185:328–40

23. Khuri SF, Henderson WG, Daley J, Jonasson O, Jones RS, Campbell DA Jr, Fink AS, Mentzer RM Jr, Steeger JE. The patient safety in surgery study: Background, study design, and patient populations. J Am Coll Surg. 2007;204:1089–102

24. AHRQ’s Ambulatory Safety and Quality Program: Health IT Portfolio. Rockville, Maryland, Agency for Healthcare Research and Quality, 2012

25. Awad IT, Chung F. Factors affecting recovery and discharge following ambulatory surgery. Can J Anaesth. 2006;53:858–72

26. Ford SJ, Wheeler JM, Borley NR. Factors influencing selection for a day-case or 23-h stay procedure in transanal endoscopic microsurgery. Br J Surg. 2010;97:410–4

27. Fleisher LA, Pasternak LR, Herbert R, Anderson GF. Inpatient hospital admission and death after outpatient surgery in elderly patients: Importance of patient and system characteristics and location of care. Arch Surg. 2004;139:67–2

28. Bryson GL, Chung F, Cox RG, Crowe MJ, Fuller J, Henderson C, Finegan BA, Friedman Z, Miller DR, van Vlymen J. Patient selection in ambulatory anesthesia—An evidence-based review: Part II. Can J Anaesth. 2004;51:782–94

29. Procter LD, Davenport DL, Bernard AC, Zwischenberger JB. General surgical operative duration is associated with increased risk-adjusted infectious complication rates and length of hospital stay. J Am Coll Surg. 2010;210:60–5.e1–2

30. . Physical status: The use and interpretation of anthropometry. Report of a WHO Expert Committee. World Health Organ Tech Rep Ser. 1995;854:1–452

31. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline for Prevention of Surgical Site Infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Am J Infect Control. 1999;27:97–132 quiz 133–4; discussion 96

32. Ercole FF, Starling CE, Chianca TC, Carneiro M. Applicability of the national nosocomial infections surveillance system risk index for the prediction of surgical site infections: A review. Braz J Infect Dis. 2007;11:134–41

33. Raval MV, Cohen ME, Ingraham AM, Dimick JB, Osborne NH, Hamilton BH, Ko CY, Hall BL. Improving American College of Surgeons National Surgical Quality Improvement Program risk adjustment: Incorporation of a novel procedure risk score. J Am Coll Surg. 2010;211:715–23

34. Whippey A, Kostandoff G, Paul J, Ma J, Thabane L, Ma HK. Predictors of unanticipated admission following ambulatory surgery: A retrospective case-control study. Can J Anaesth. 2013;60:675–83

35. Harrell FE Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis, Springer Series in Statistics. 2001 New York Springer:90–2

36. Aronson WL, McAuliffe MS, Miller K. Variability in the American Society of Anesthesiologists Physical Status Classification Scale. AANA J. 2003;71:265–74

Back to Top | Article Outline
Appendix 1.
Back to Top | Article Outline
Appendix 2.
NSQIP Data Fields Us...
NSQIP Data Fields Us...
Image Tools
Back to Top | Article Outline
Appendix 3.
Predictor VariablesN...
Predictor VariablesN...
Image Tools
Primary Outcome Vari...
Primary Outcome Vari...
Image Tools

© 2013 American Society of Anesthesiologists, Inc.

Publication of an advertisement in Anesthesiology Online does not constitute endorsement by the American Society of Anesthesiologists, Inc. or Lippincott Williams & Wilkins, Inc. of the product or service being advertised.

Article Tools