- Email: [email protected]
Hospital admission following ambulatory surgery
Hospital Admission following Ambulatory Surgery A. G. Greenburg,
MD, PhD, Joan P. Greenburg, MD, Annie Tewel, MD, Christopher Odalys Machin, MD, Simon McRae, MD, Providence, Rhode /s/and
BACKGROUND: Ambulatory surgery continues to grow in quantity and complexity of procedures. Effective measures of “quality” are not readily apparent. “Unplanned admission rate” may well reflect the quality of care in this area. Identifying factors related to this event could be helpful in quality assessment and improvement. METHODS: A review of all unplanned admissions for a 3-year period in a University-affiliated teaching hospital. RESULTS: An overall rate of 0.85% (129/15,132) was observed. Rate varies by specialty and no one procedure was at higher risk. Pain control, cardiopulmonary, and bleeding problems as well as larger than anticipated procedures accounted for 73% of the admissions. CONCLUSIONS: Unplanned admission following ambulatory surgery is relatively rare but could reflect overall quality in terms of the system, physician, and patient. Comparisons between institutions and within institution requires defining key demographic elements whose identification for now remains a challenge. Am J Surg. 1996;172:21-23.
I
n the current era 50% to 60% of all surgical procedures are being performed on an “out-patient” basis. This spectacular growth and development of ambulatory surgery cannot go unnoticed. Moreover, the rapid and significant developments in “minimally invasive surgery” by many surgical specialties will add volutne in the future. Few of our surgical forefathers would have predicted the current status and would stand in awe of what we accomplish today in the outpatient setting. Many aspects of ambulatory surgery could be discussed. Included would be preoperative assessment, anesthesia and analgesia, patient management, surgical technique, support systems, facility design, and quality assurance. Taken together, these elements must interact to assure that patient safety and quality care are exercised at the highest levels. Additional concerns include the decision to perform surgery at an outpatient/ambulatory site and overall patient risk assessment. Overriding these issues an d concerns is the defmition and implementation of a From the Department of Surgery/The Miriam Hospital, Brown University School of Medicine, Providence, Rhode Island. Requests for reprints should be addressed to A.G. Greenburg University
School
C 1996 by Excerpta All rights reserved.
of Medicine,
Medica,
164 Summit
Inc.
Avenue,
Providence,
Breen, MD,
quality assurance program for ambulatory surgery. Without a postoperative period of hosplt,llization it may be difficult to ascertain traditIona morbidity rates because these patients are at home and surveillance is less acute or intense. Morbidity and mortality rates still reflect the quality of care provided by a system, institution, or individual physicians especially when properly adjusted for risk. Identification of populaticms at risk is, of course, an important element. These traditional markers, morbidity and mortality, may not be appn,priate measures of quality for ambulatory surgery. Mortality should he a relati\:cly slnall if not minimal number; ‘,’ tnorhidity data may he difticult to obtain for any number of reasons. With pressures for ongoing healthcare reform driven primarily by cost issues there is a need to assure the public and the payers that ambulatory surgery is safe, effective, and ultimately economical. A quality assurance program that assesses the surgeon and the facility, stratified by individual patient risk, is required to achieve this end. The unplanned admission rate has been one‘ measure of the quality of ambulatory surgery. Taken in Its broadest context, this measure may, in fact, reflect quality. This is a complex variable made up of elements representing individually and in combination the patient, the physician, and the facility, each on many levels. This report, descriptive in nature, explores some of these v,lriables related to unplanned postoperative hospital admission following ambulatory surgery. Emphasis is on identifying some prohahle patient risk factors. This is a start at identifying eletnents that could influence the unplanned admission rate following ambulatory surgery. Given that specific factors can he identified it is apparent that methods to correct the underlying causes could he implemented.
MATERIALS
AND
METHODS
From the operative-log data base system of The Miriam Hospital all operations performed as outpatients in our Surgicenter were selected. The time from October 1, 1991 through September 30, 1994 (3 fiscal years) was chosen. Only patients scheduled on whom surgery was performed as an outpatient were selected. Cases admitted following the surgical procedure were obtained from a log maintained in the Surgicenter recovery area. Data collected included service, physician/surgeon, patient age and gender, anesthesia type, ASA classification, reason for admission, duration of procedure and time of day the procedure started. During this period 15,132 patient operations were performed in the Surgicenter. The vast tnajority were general surgical or surgical specialty cases. There were 129 unplan0002-961 O/96/$1 PII SOOO2-9610(96)00050-5
5.00
21
OSPITAL
ADMISSION
FOLLOWING
SURGERY/GREENBUl?G
ET AL
DISCUSSION
ned admissions to the hospital. Data from this group of patients was the subject of further analysis to determine the overall incidence of problems and to identify care patterns that may be corrected.
RESULTS The overall unanticipated admission rate was 0.85% ( 129/ 15,132). The mean age of these cases was 56.4 z? 17.6 years with a mean ASA classification of 2.62 + 0.8. There were 55 men (43%) and 74 women (57%); our ambulatory population is about 1 male:2 female overall from a random sample of an additional 200 patients. The distribution of cases by service is shown in Table I. Admissions following general surgical procedures are out of proportion to the volume of general surgical cases and comprise the largest number and percentage of admissions. These 75 patients, with a 1.42% admission rate, represent the gamut of general surgery and no specific procedure can be identified as primarily responsible. Orthopedics; ear, nose, and throat (ENT); and neurosurgery also have a slightly higher admission rate compared to that expected from the case distribution; again, no one procedure was responsible. The primary reason for admission is shown in Table II. Some patients had multiple reasons, however, the primary one listed on the log was selected as “related” and used for tabulation. Overall, there were 147 reasons stated for the 129 admitted patients; 16 patients had 2 or 3 reasons cited. The requirement for pain medication led to 24.8% of the admissions. Cardiopulmonary problems were implicated in 2 1% of the cases. Coagulation, bleeding or wound problems accounted for 14.7% of the admissions. Procedure-related events, usually a larger procedure than anticipated, led to 13.9% of the admissions. Nausea and vomiting, presumed to be related to the use of anesthesia or intraoperative analgesia, accounted for 11.6% of the admitted cases. The remainder were divided equally, 4.7% each, between the need for intravenous antibiotics, electrolyte problems including hypoglycemia and urinary retention. Three services accounted for 86.8% of the admissions: general surgery, ophthalmology, and orthopedics. Overall, these three services accounted for only 75% of the volume in the Surgicenter in this time period. TABLE
I Distribution
Service:
of Cases
Cases
General surgery Opthalmology Orthopedics* Plastic surgery Urology ENT Gynecology Neurosurgery Other**
5295 4091 2005 1260 1255 516 81 50 591
by Service
% Cases
Unplanned Unplanned Admissions
35.00 27.00 13.30 8.33 a.29 3.40 0.53 0.33 3.90
Admissions % Rate
75 17 20 5 3 6 0 1 2
1.42 0.42 0.99 0.40 0.24 1.16 0 2.00 0.34
* Including: Podiatry, hand ** Including: Thoracic, anesthesia, endoscopy under general anesthesfa EAIT: ear, nose, and throat Note: co/. 2 total # cases = 15,144, not the stated 75,732 (in ted) but % cases
22
THE
AMERICAN
JOURNAL
OF SURGERY@
VOLUME
172
are
The series reported here may be one of the first systematic reporting of reasons for unplanned admission following ambulatory surgery. As such it represents a reference frame upon which a quality assurance program could be built. Inherent in that program is a rational basis for identifying problemspatient, physician, system-that can be investigated and resolved. Unplanned admissions are significant events and should be investigated and root causes identified and addressed if overall quality is to be maintained. The overall rate reported here, 0.85%, is well within the ranges reported in the literature. ‘M Of note, the two lowest rates, 0.28”/ and 0.7%, were from Canada and England, respectively where health care financing is quite different. The overall range of 0.28% to 1.34% is a generally accepted guideline for the United States. Direct comparisons are, of course, difficult for a variety of reasons. Clearly there are “systems issues” which are obvious. In these studies the populations are not stratified by service only by the facility. Guidelines for patient selection are not apparent yet necessary if one is to effect real comparisons. Patient selection criteria are critical elements in the process to assure both safety and quality. Age did not appear to be a factor in the population admitted. This could be explamed by the use of defined and planned preprocedure testing that seeks significant co-morbidities and addresses them in an anticipatory fashion’,” The use of multiple screens-nurses at preprocedure screening, anesthesia review, nurses on admission-may prevent patients from slipping through who are otherwise at significant risk. Thirty percent (N = 39) of the patients were over age 70, with 11 older than 80 years. Just under 40% of the patients were 49 years old or younger (N = 49). It would appear that the reasons for admission is independent of age unless one postulates pain relief is more difficult in the young or the old, which does not appear to be the situation. On the other hand, over 60% of the cardiac problems occurred in the patients 70 years or older, indicating the potential risk of this group for cardiopulmonary events. The 18 procedure-related cases require comment. In these it could be the disease process was more extensive than originally considered and a larger procedure was indicated. Whether this could have been avoided becomes the quality issue. It would be necessary to evaluate the diagnostic studies and the decision making process of the surgeon to determine why this came about. Clearly it is % Cases important and useful information for the Admitted Surgicenter when planning and at58.14 tempting to maintain function if the 13.18 procedure time can be properly esti15.50 mated. A procedure that does not go as 3.88 planned, and this may happen, has the 2.33 “domino effect” of delaying all subse4.65 quent cases. This delay may be per0 ceived as a facility quality issue when in 0.78 1.55 fact it is unavoidable. No specific pattern by surgeon was apparent in the cases that were deemed to be “larger than anticipated.” Of note, six of the correct to 75,132 seven of these were in general surgery but covered the spectrum of groin and
JULY
1996
[HOSPITAL TABLE
II Primary Reason for Admission Ambulatory Surgery N
Pain Cardiac/Pulmonary Arrhythmia BP Change Syncope Congestive Failure R/O Ml Pulmonary Failure Bleeding/Wounds Procedure Related* Nausea/Vomiting Requires IV Antibiotics Electrolyte Problems Hypoglycemia K+ Urinary Retention + Longer
than arkipated;
following % of Admissions
32 27 11 4 3 3 4 2 19 18 15 6 6 4 2 6 need
for observation;
24.8 21 .o
14.7 13.9 11.6 4.7 4.7
4.7 appliance;
unable
to D/C.
anorectal procedures primarily. These tended to be at the extremes of age, both young and old. Three of these cases resulted from complications of central line placement; all of these patients subsequently were discharged from the hospital alive. Congenital bleeding disorders are unusual in this population because of the preprocedure screening. Patients who have bleeding or coagulopathy problems in the perioperative period are usually taking nonsteroidal anti-inflammatory agents. In the preprocedure assessment emphasis on stopping these should be exercised and the elective procedure delayed if necessary until the drug effects have abated. Good patient instructions can obviate many of these problems. Similarly, by preoperatively adjusting the insuiin dose in patients with insulin-dependent diabetes, hyperglycemia requiring admission may be an avoidable issue. The problem of nausea and vomiting following general anesthesia can be managed through dosing and appropriate anesthetic management techniques. Matching the anesthe-
THE
~13Mlss10~
FOLLOWING
SURGERY~GREENBURG
ET AL
sia to the procedure is important as ion:: as patient safety is paramount and the surgical technique is not compromised.’ The ASA classification requires comment. Although it is universally used as a “risk assessment tool” it was not designed for use in the ambulatory environment. Indeed, as technology progresses more and more patients at higher ASA levels are being done in the ambulatory environment. The key to success is proper patlent selection and appropriate support systems.’ Unplanned admission following ambulatory surgery is a relatively uncommon event. It can he a marker of overall system quality and followed for patterns to detect deviations based on patient, physician, or system issues. Because it is a relatively unusual event our data nray be skewed as it represents one population and demographic set that may not be easily extrapolated to another setting or population. The uniqueness of an organization, its practice, its patient demography, and capabilities clearly limit direct comparison between units. Unless these aspects are tnade explicit it will in fact hc difficult to compare Licilities for quality. All of these elements must be taken together to create a system and mechanism that measures quality and assures patient safety. For now that remains a challenge.
REFERENCES 1. Natof HE. Complicarions aastxiattxl wth amhulator\i surgerv. /AMA. 198@;244:1il6-1118. 2. Warner MA, Shirlds SE, Chute CG. M,IICI~ mothidity and mortality wthin one mmth of ambulatory \nrgrry dnd anesthesia. J.4M.4. 1993;270:1437-1441. 3. johnson CL>, jxrett PE. Admission ro hwpital after day care surgery. Ann Royl Cal Surg En,o. 1990;7?:225-228. 4. Fancourt-Smith I’F, Hornstem J, Jenkins LC. IHospital admissions from the surgical day care center of Vmcouver General Hospital 1977-1987. CanJ Anacsth. 1990;37:69%7O4. 5. Biswar TK, Leary C. Postoperative h~qoral ,rdrnlsslcm from a day surgery unit: a xl-en year retrospectwe >urvey. Anuesth Intensive care 1992;20:147-150. 6. Oshotne GA, Rudkin GE. 0utcome aft<~t- day surgery in a major teaching hospiral. .4naesth Intenswe (2~ lW1;?1:822-827. 7. Greenburg A(;. Ambulatory surgery in rhr ‘30’s. Ambulatory Surg 1994;2:136&141. 8. Golub K, Canru R, Sortento JJ, Stein tlD. Efficacy of pre-admission test& in ambulatory surgical patxnts. Am J Surg. 163:5657-571.
AMERICAN
JOURNAL
OF SURGERY@
VOLUME
172
JULY
1996
23
MD, PhD, Joan P. Greenburg, MD, Annie Tewel, MD, Christopher Odalys Machin, MD, Simon McRae, MD, Providence, Rhode /s/and
BACKGROUND: Ambulatory surgery continues to grow in quantity and complexity of procedures. Effective measures of “quality” are not readily apparent. “Unplanned admission rate” may well reflect the quality of care in this area. Identifying factors related to this event could be helpful in quality assessment and improvement. METHODS: A review of all unplanned admissions for a 3-year period in a University-affiliated teaching hospital. RESULTS: An overall rate of 0.85% (129/15,132) was observed. Rate varies by specialty and no one procedure was at higher risk. Pain control, cardiopulmonary, and bleeding problems as well as larger than anticipated procedures accounted for 73% of the admissions. CONCLUSIONS: Unplanned admission following ambulatory surgery is relatively rare but could reflect overall quality in terms of the system, physician, and patient. Comparisons between institutions and within institution requires defining key demographic elements whose identification for now remains a challenge. Am J Surg. 1996;172:21-23.
I
n the current era 50% to 60% of all surgical procedures are being performed on an “out-patient” basis. This spectacular growth and development of ambulatory surgery cannot go unnoticed. Moreover, the rapid and significant developments in “minimally invasive surgery” by many surgical specialties will add volutne in the future. Few of our surgical forefathers would have predicted the current status and would stand in awe of what we accomplish today in the outpatient setting. Many aspects of ambulatory surgery could be discussed. Included would be preoperative assessment, anesthesia and analgesia, patient management, surgical technique, support systems, facility design, and quality assurance. Taken together, these elements must interact to assure that patient safety and quality care are exercised at the highest levels. Additional concerns include the decision to perform surgery at an outpatient/ambulatory site and overall patient risk assessment. Overriding these issues an d concerns is the defmition and implementation of a From the Department of Surgery/The Miriam Hospital, Brown University School of Medicine, Providence, Rhode Island. Requests for reprints should be addressed to A.G. Greenburg University
School
C 1996 by Excerpta All rights reserved.
of Medicine,
Medica,
164 Summit
Inc.
Avenue,
Providence,
Breen, MD,
quality assurance program for ambulatory surgery. Without a postoperative period of hosplt,llization it may be difficult to ascertain traditIona morbidity rates because these patients are at home and surveillance is less acute or intense. Morbidity and mortality rates still reflect the quality of care provided by a system, institution, or individual physicians especially when properly adjusted for risk. Identification of populaticms at risk is, of course, an important element. These traditional markers, morbidity and mortality, may not be appn,priate measures of quality for ambulatory surgery. Mortality should he a relati\:cly slnall if not minimal number; ‘,’ tnorhidity data may he difticult to obtain for any number of reasons. With pressures for ongoing healthcare reform driven primarily by cost issues there is a need to assure the public and the payers that ambulatory surgery is safe, effective, and ultimately economical. A quality assurance program that assesses the surgeon and the facility, stratified by individual patient risk, is required to achieve this end. The unplanned admission rate has been one‘ measure of the quality of ambulatory surgery. Taken in Its broadest context, this measure may, in fact, reflect quality. This is a complex variable made up of elements representing individually and in combination the patient, the physician, and the facility, each on many levels. This report, descriptive in nature, explores some of these v,lriables related to unplanned postoperative hospital admission following ambulatory surgery. Emphasis is on identifying some prohahle patient risk factors. This is a start at identifying eletnents that could influence the unplanned admission rate following ambulatory surgery. Given that specific factors can he identified it is apparent that methods to correct the underlying causes could he implemented.
MATERIALS
AND
METHODS
From the operative-log data base system of The Miriam Hospital all operations performed as outpatients in our Surgicenter were selected. The time from October 1, 1991 through September 30, 1994 (3 fiscal years) was chosen. Only patients scheduled on whom surgery was performed as an outpatient were selected. Cases admitted following the surgical procedure were obtained from a log maintained in the Surgicenter recovery area. Data collected included service, physician/surgeon, patient age and gender, anesthesia type, ASA classification, reason for admission, duration of procedure and time of day the procedure started. During this period 15,132 patient operations were performed in the Surgicenter. The vast tnajority were general surgical or surgical specialty cases. There were 129 unplan0002-961 O/96/$1 PII SOOO2-9610(96)00050-5
5.00
21
OSPITAL
ADMISSION
FOLLOWING
SURGERY/GREENBUl?G
ET AL
DISCUSSION
ned admissions to the hospital. Data from this group of patients was the subject of further analysis to determine the overall incidence of problems and to identify care patterns that may be corrected.
RESULTS The overall unanticipated admission rate was 0.85% ( 129/ 15,132). The mean age of these cases was 56.4 z? 17.6 years with a mean ASA classification of 2.62 + 0.8. There were 55 men (43%) and 74 women (57%); our ambulatory population is about 1 male:2 female overall from a random sample of an additional 200 patients. The distribution of cases by service is shown in Table I. Admissions following general surgical procedures are out of proportion to the volume of general surgical cases and comprise the largest number and percentage of admissions. These 75 patients, with a 1.42% admission rate, represent the gamut of general surgery and no specific procedure can be identified as primarily responsible. Orthopedics; ear, nose, and throat (ENT); and neurosurgery also have a slightly higher admission rate compared to that expected from the case distribution; again, no one procedure was responsible. The primary reason for admission is shown in Table II. Some patients had multiple reasons, however, the primary one listed on the log was selected as “related” and used for tabulation. Overall, there were 147 reasons stated for the 129 admitted patients; 16 patients had 2 or 3 reasons cited. The requirement for pain medication led to 24.8% of the admissions. Cardiopulmonary problems were implicated in 2 1% of the cases. Coagulation, bleeding or wound problems accounted for 14.7% of the admissions. Procedure-related events, usually a larger procedure than anticipated, led to 13.9% of the admissions. Nausea and vomiting, presumed to be related to the use of anesthesia or intraoperative analgesia, accounted for 11.6% of the admitted cases. The remainder were divided equally, 4.7% each, between the need for intravenous antibiotics, electrolyte problems including hypoglycemia and urinary retention. Three services accounted for 86.8% of the admissions: general surgery, ophthalmology, and orthopedics. Overall, these three services accounted for only 75% of the volume in the Surgicenter in this time period. TABLE
I Distribution
Service:
of Cases
Cases
General surgery Opthalmology Orthopedics* Plastic surgery Urology ENT Gynecology Neurosurgery Other**
5295 4091 2005 1260 1255 516 81 50 591
by Service
% Cases
Unplanned Unplanned Admissions
35.00 27.00 13.30 8.33 a.29 3.40 0.53 0.33 3.90
Admissions % Rate
75 17 20 5 3 6 0 1 2
1.42 0.42 0.99 0.40 0.24 1.16 0 2.00 0.34
* Including: Podiatry, hand ** Including: Thoracic, anesthesia, endoscopy under general anesthesfa EAIT: ear, nose, and throat Note: co/. 2 total # cases = 15,144, not the stated 75,732 (in ted) but % cases
22
THE
AMERICAN
JOURNAL
OF SURGERY@
VOLUME
172
are
The series reported here may be one of the first systematic reporting of reasons for unplanned admission following ambulatory surgery. As such it represents a reference frame upon which a quality assurance program could be built. Inherent in that program is a rational basis for identifying problemspatient, physician, system-that can be investigated and resolved. Unplanned admissions are significant events and should be investigated and root causes identified and addressed if overall quality is to be maintained. The overall rate reported here, 0.85%, is well within the ranges reported in the literature. ‘M Of note, the two lowest rates, 0.28”/ and 0.7%, were from Canada and England, respectively where health care financing is quite different. The overall range of 0.28% to 1.34% is a generally accepted guideline for the United States. Direct comparisons are, of course, difficult for a variety of reasons. Clearly there are “systems issues” which are obvious. In these studies the populations are not stratified by service only by the facility. Guidelines for patient selection are not apparent yet necessary if one is to effect real comparisons. Patient selection criteria are critical elements in the process to assure both safety and quality. Age did not appear to be a factor in the population admitted. This could be explamed by the use of defined and planned preprocedure testing that seeks significant co-morbidities and addresses them in an anticipatory fashion’,” The use of multiple screens-nurses at preprocedure screening, anesthesia review, nurses on admission-may prevent patients from slipping through who are otherwise at significant risk. Thirty percent (N = 39) of the patients were over age 70, with 11 older than 80 years. Just under 40% of the patients were 49 years old or younger (N = 49). It would appear that the reasons for admission is independent of age unless one postulates pain relief is more difficult in the young or the old, which does not appear to be the situation. On the other hand, over 60% of the cardiac problems occurred in the patients 70 years or older, indicating the potential risk of this group for cardiopulmonary events. The 18 procedure-related cases require comment. In these it could be the disease process was more extensive than originally considered and a larger procedure was indicated. Whether this could have been avoided becomes the quality issue. It would be necessary to evaluate the diagnostic studies and the decision making process of the surgeon to determine why this came about. Clearly it is % Cases important and useful information for the Admitted Surgicenter when planning and at58.14 tempting to maintain function if the 13.18 procedure time can be properly esti15.50 mated. A procedure that does not go as 3.88 planned, and this may happen, has the 2.33 “domino effect” of delaying all subse4.65 quent cases. This delay may be per0 ceived as a facility quality issue when in 0.78 1.55 fact it is unavoidable. No specific pattern by surgeon was apparent in the cases that were deemed to be “larger than anticipated.” Of note, six of the correct to 75,132 seven of these were in general surgery but covered the spectrum of groin and
JULY
1996
[HOSPITAL TABLE
II Primary Reason for Admission Ambulatory Surgery N
Pain Cardiac/Pulmonary Arrhythmia BP Change Syncope Congestive Failure R/O Ml Pulmonary Failure Bleeding/Wounds Procedure Related* Nausea/Vomiting Requires IV Antibiotics Electrolyte Problems Hypoglycemia K+ Urinary Retention + Longer
than arkipated;
following % of Admissions
32 27 11 4 3 3 4 2 19 18 15 6 6 4 2 6 need
for observation;
24.8 21 .o
14.7 13.9 11.6 4.7 4.7
4.7 appliance;
unable
to D/C.
anorectal procedures primarily. These tended to be at the extremes of age, both young and old. Three of these cases resulted from complications of central line placement; all of these patients subsequently were discharged from the hospital alive. Congenital bleeding disorders are unusual in this population because of the preprocedure screening. Patients who have bleeding or coagulopathy problems in the perioperative period are usually taking nonsteroidal anti-inflammatory agents. In the preprocedure assessment emphasis on stopping these should be exercised and the elective procedure delayed if necessary until the drug effects have abated. Good patient instructions can obviate many of these problems. Similarly, by preoperatively adjusting the insuiin dose in patients with insulin-dependent diabetes, hyperglycemia requiring admission may be an avoidable issue. The problem of nausea and vomiting following general anesthesia can be managed through dosing and appropriate anesthetic management techniques. Matching the anesthe-
THE
~13Mlss10~
FOLLOWING
SURGERY~GREENBURG
ET AL
sia to the procedure is important as ion:: as patient safety is paramount and the surgical technique is not compromised.’ The ASA classification requires comment. Although it is universally used as a “risk assessment tool” it was not designed for use in the ambulatory environment. Indeed, as technology progresses more and more patients at higher ASA levels are being done in the ambulatory environment. The key to success is proper patlent selection and appropriate support systems.’ Unplanned admission following ambulatory surgery is a relatively uncommon event. It can he a marker of overall system quality and followed for patterns to detect deviations based on patient, physician, or system issues. Because it is a relatively unusual event our data nray be skewed as it represents one population and demographic set that may not be easily extrapolated to another setting or population. The uniqueness of an organization, its practice, its patient demography, and capabilities clearly limit direct comparison between units. Unless these aspects are tnade explicit it will in fact hc difficult to compare Licilities for quality. All of these elements must be taken together to create a system and mechanism that measures quality and assures patient safety. For now that remains a challenge.
REFERENCES 1. Natof HE. Complicarions aastxiattxl wth amhulator\i surgerv. /AMA. 198@;244:1il6-1118. 2. Warner MA, Shirlds SE, Chute CG. M,IICI~ mothidity and mortality wthin one mmth of ambulatory \nrgrry dnd anesthesia. J.4M.4. 1993;270:1437-1441. 3. johnson CL>, jxrett PE. Admission ro hwpital after day care surgery. Ann Royl Cal Surg En,o. 1990;7?:225-228. 4. Fancourt-Smith I’F, Hornstem J, Jenkins LC. IHospital admissions from the surgical day care center of Vmcouver General Hospital 1977-1987. CanJ Anacsth. 1990;37:69%7O4. 5. Biswar TK, Leary C. Postoperative h~qoral ,rdrnlsslcm from a day surgery unit: a xl-en year retrospectwe >urvey. Anuesth Intensive care 1992;20:147-150. 6. Oshotne GA, Rudkin GE. 0utcome aft<~t- day surgery in a major teaching hospiral. .4naesth Intenswe (2~ lW1;?1:822-827. 7. Greenburg A(;. Ambulatory surgery in rhr ‘30’s. Ambulatory Surg 1994;2:136&141. 8. Golub K, Canru R, Sortento JJ, Stein tlD. Efficacy of pre-admission test& in ambulatory surgical patxnts. Am J Surg. 163:5657-571.
AMERICAN
JOURNAL
OF SURGERY@
VOLUME
172
JULY
1996
23