Unanticipated Hospital Admission After Ambulatory Surgery: The Devil Is in the Details

See Article, p 497 Improvements in perioperative care and a changing pattern of health care payments have resulted in a significant expansion of procedures performed on an outpatient basis. This growth is likely to continue, as more complex and invasive surgeries are increasingly being performed in the ambulatory setting. Similarly, complex noninvasive procedures performed in nonoperating room settings―for example, in gastroenterology endoscopy, interventional cardiovascular, and interventional radiology suites―have dramatically risen. While the proliferation of day cases has several benefits, including reduced health care costs,1 it is critical that the migration of procedures from an inpatient to outpatient setting does not compromise patient safety and quality of care. The US Center for Medicare and Medicaid Services (CMS) considers unplanned transfers directly from an ambulatory surgery facility to an acute care hospital as one of the quality measures for such ambulatory facilities.2 The other quality metrics for ambulatory facilities can include subsequent acute care encounters, emergency department visits, and hospital admissions after initial discharge home. Recently, CMS introduced any admission within 7 days after day case colonoscopy as a quality metric.2 It is expected that this metric will be extended to other procedures. Unplanned hospital transfer and admission rates after discharge home are also used to determine which procedures and patients are suitable for ambulatory surgery.3–7 However, the literature on these outcome measures is sparse. Using large databases from 3 tertiary care, academic hospitals in Massachusetts and from the Healthcare Cost and Utilization Project (HCUP) in New York state, Teja et al8 evaluated the incidence and reasons for unplanned hospital admission in the patients who were discharged home within 24 hours after a procedure. These investigators are to be lauded for undertaking a trial of this magnitude and rigor. A major strength of this study is that the analyses included data from academic and nonacademic practices, as well as data from hospital-based and nonhospital-based ambulatory facilities. Furthermore, a diverse patient population (ie, those with Medicare and Medicaid, as well as private health care insurance) and a wide range of procedures (ie, nonoperating room procedures and surgical procedures) were included. This allows the conclusions from the study to be generalized to a broader group of patients and procedures. The information derived from this study could also be used to determine the suitability of procedures and patients in a day care setting. Teja et al8 also identified several comorbid conditions associated with high unplanned admission rate―namely, chronic pulmonary disease, congestive heart failure, diabetes mellitus, renal failure, liver disease, peripheral vascular disease, deficiency anemia, drug abuse, and depression. These medical conditions could serve as indicators for focused patient evaluation and optimization before the day of surgery. Optimization of these comorbid conditions should improve outcomes after ambulatory surgery.9 The rapid migration of more extensive surgical procedures and medically complex patients to the ambulatory setting has created uncertainty among anesthesiologists who must assess their suitability and maintain high-quality and safe perioperative care. Patient selection is a multifaceted and dynamic process that depends on an interplay among the surgical procedure, patient clinical characteristics, and the anesthetic technique (sedation/analgesia, local/regional anesthesia, or general anesthesia). For example, procedures like cataract surgery, which are commonly performed under topical anesthesia with minimal or no sedation, could be performed in a free-standing ambulatory surgery facility irrespective of the patient characteristics like age and comorbid conditions. In contrast, patient characteristics and comorbid conditions would play a key role in determining suitability for more extensive surgical procedures that require general anesthesia. Patient selection also depends on the type of ambulatory setting, specifically, a hospital outpatient department (HOPD), free-standing ambulatory surgery center (ASC), or office-based surgery. Unlike a free-standing ASC, an HOPD, which is physically connected to a hospital or located within close proximity of the hospital, has the ability to manage medically complex patients undergoing more extensive procedures because of the immediate availability of specialists for consultation, advanced equipment, and ancillary services, like laboratory, blood bank, and respiratory therapy. Similarly, nonoperating settings in a hospital may be able to perform more invasive procedures in sicker and older patients, which may not be possible in a free-standing ASC. Teja et al8 observed that most of the surgical procedures with the highest readmission rates included emergency surgery and more invasive ones like mediastinoscopy, thyroidectomy, and abdominal hysterectomy. Such procedures are more likely to be performed in HOPD or an ambulatory surgery facility with the ability to provide an extended recovery stay10—not in a typical free-standing ASC. Similarly, the procedures in the nonoperating room setting with highest unplanned hospital admission rates included endoscopic retrograde cholangiopancreatography, bronchoscopy and placement of bronchial stents, central venous access and insertion of intravascular vena cava filter, thoracentesis, abdominal paracentesis, and intraperitoneal catheterization, which again are commonly performed in an HOPD setting. These observations suggest that current clinical practice in an HOPD setting needs closer scrutiny. On the other hand, clinical care in the freestanding ASC setting appears to be appropriate. However, this distinction was not assessed by Teja et al.8 Teja et al8 also developed a prediction tool (PREdicting admission after Outpatient Procedures [PREOP]) that could be used for quality improvement in and benchmarking of ambulatory facilities. This novel scoring system can also be used to determine procedure suitability and patient eligibility for the ambulatory setting. For example, a PREOP score of 14 corresponds to a 30-day unplanned admission rate of 2%, which was the mean value observed in this study. The PREOP score could thus be used to guide procedure and patient selection. Specifically, for example, procedures with a PREOP cut point score of 14 or greater could be performed on an ambulatory basis only if the patient is healthy. Similarly, certain emergent/urgent surgical procedures may be suitable for an HOPD setting, assuming that the patients are relatively healthy. In contrast, surgical procedures with PREOP scores of greater than 14 (those involving respiratory system, mediastinum and diaphragm, and vascular system) may not be suitable for ambulatory surgery even in healthy patients. The findings of Teja et al8 should be considered in the context of certain limitations, some of which the authors acknowledge. It is well known that administrative databases have inherent inconsistencies in coding of the identifiers. More importantly, such retrospective analyses may not continue to be relevant in the current rapidly changing ambulatory practice environment. Furthermore, the databases used in this study did not allow for identification of postdischarge acute care or emergency department visits without subsequent hospitalization. Teja et al8 also did not include the rate of direct acute care hospital transfer, which remains a measure used by CMS to assess quality of care at an ambulatory facility. However, such patient transfer should not be used as an independent measure of adverse outcomes in an ambulatory setting.11 Moreover, unplanned admission on the day of surgery or within 24 hours after surgery was not assessed, a metric that could be used to determine whether a specific patient undergoing a specific procedure should be admitted overnight. Importantly, the validity of 30-day unplanned admission rates as a quality measure for ambulatory setting can be criticized, because the setting where the procedure is performed (outpatient versus inpatient) should have little influence on outcomes occurring several weeks later. The lack of specific clinical information like severity of comorbid conditions and anesthetic type may also limit the utility of the findings of Teja et al8 in assisting with patient selection on a routine basis. Given the aforementioned complexities in appropriate selection of the ambulatory setting to achieve best outcomes, it would have been helpful if the data from HOPD settings and freestanding ASC settings were analyzed separately. Similarly, it would have been beneficial if the data from nonoperating room settings were separated from the surgical setting. In summary, the unplanned admission rates after ambulatory procedures remain low despite the liberalizing of procedure suitability and patient eligibility. The proposed PREOP scoring system is an important step toward improving quality of care in the ambulatory setting; however, it needs to be prospectively validated. The procedure-related predictors of adverse outcomes could be used to determine the optimal ambulatory setting (HOPD versus ASC). Similarly, the patient-related predictors of adverse outcomes should allow clinicians to modify their perioperative care via formal preoperative optimization of high-risk patients. Future studies are needed to confirm if modification of perioperative care, such as triaging of high-risk patients for preoperative optimization, as well as the targeted use of postoperative interventions, improve meaningful postoperative outcomes. Future studies should explore unplanned readmissions during the immediate 24–48 hours after day care procedures. Future studies should also compare procedure-specific 7- and 30-day outcomes for the various settings (inpatient hospital stay versus HOPD versus ASC) while controlling for patient characteristics. Finally, development of procedure-specific selection criteria for ambulatory surgery is crucial so that it can be easily integrated in an applicable enhanced recovery pathway. DISCLOSURES Name: Girish P. Joshi, MBBS, MD, FFARCSI. Contribution: This author helped write and revise the manuscript. Conflicts of Interest: G. P. Joshi reported honoraria from Baxter Pharmaceuticals and Pacira Pharmaceuticals. Name: Thomas R. Vetter, MD, MPH. Contribution: This author helped write and revise the manuscript. Conflicts of Interest: None. This manuscript was handled by: Jean-Francois Pittet, MD.