The Way We See It: Surgeons' Perceptions of Perioperative Events

Abstract

Objectives: The objective of this study was to examine gynecologic (GYN) surgeons' perceptions of perioperative events and their severity. Secondarily, differences in perceptions between surgeons performing elective and nonelective surgery were compared. Design: Fifty-seven perioperative scenarios associated with GYN surgery were created. The scenarios were grouped into 16 broader categories (e.g., vascular injury, neuropathy). Scenarios were electronically distributed to GYN surgeons from geographically diverse areas, and surgeons classified each scenario as either a complication or a routine perioperative event. Surgeons rated complication severity using both a numerical rating scale (1–100) and the Clavien–Dindo scale (I–V). Study participants were divided into two groups based on whether they performed elective or nonelective surgeries. Materials and Methods: PASW^® version 18 was used to calculate means, standard deviations, intraclass correlation coefficients (ICC), and coefficients of variability (Cv) for severity rankings of each scenario within and among surgeon groups. Results: Sixty-three out of 82 surgeons (77%) responded. GYN oncologists classified conversion from laparoscopy to laparotomy, blood transfusion, and delayed return of bowel function as complications less frequently than did surgeons who performed elective surgery (36% vs. 12%, p=0.02; 62 vs. 29%, p=0.003; 85% vs. 57%, p=0.002). Urogynecologists rated postoperative urinary symptoms as more severe complications than did other surgeon groups. All groups showed high agreement (96%–100%) classifying postoperative events involving foreign body (mesh, suture) as complications (Cv=46[30–51]). Conclusions: Surgeons' perceptions of perioperative events vary based on surgical indication and subspecialty training. Future studies assessing agreement between patients and surgeons' perceptions may aid in creating a patient-oriented GYN complication assessment tool. (J GYNECOL SURG 28:270)

Introduction

Assessment of clinically important adverse events in gynecologic surgery remains a challenge. Patient preparedness and expectations surrounding the surgical experience are often the primary determinants of patient satisfaction after surgery.^1,2 However, despite a thorough preoperative informed consent discussion, patients often perceive perioperative medical events that are thought of as routine by surgeons, to be bothersome complications.¹ For example, Kenton et al. prospectively followed a cohort of women undergoing reconstructive pelvic surgery, and at 3 months after surgery asked them to list all the perioperative complications that they had encountered during the surgical process.¹ Patient responses were then compared to those listed by surgeons. The authors found significant discrepancies among patients' and surgeons' list of complications: whereas surgeons identified only 8% of the cohort as having had a postoperative complication, 44% of patients reported having experienced a complication. Such differences in patient and physician perceptions are shifting the medical community's focus toward aligning physicians' and patients' expectations for the outcomes of surgical treatment.¹ To effectively set patient expectations regarding surgery, surgeons must have clear definitions of perioperative complications, in addition to reliable ways to measure and compare outcome data.^3,4

In an attempt to standardize adverse perioperative events, Dindo and Clavien introduced a surgical complications rating system to facilitate comparison of outcomes. The Clavien–Dindo scale has become a widely used surgical complication scale, used in randomized trials and comparative cohorts in a variety of surgical specialties.^3,4 However, neither the Clavien–Dindo scale nor other currently used validated surgical complication scales capture the magnitude or consequence of complications for women undergoing gynecologic surgery. Moreover, the Clavien–Dindo scale was not originally tested to determine whether it correlates with patients' perceptions of complications.³ In a follow-up analysis, Clavien et al. found important differences in physician and patient perceptions of complication severity when they asked physicians and patients to score the severity of 30 hypothetical scenarios: patients perceived many complications as more severe than did physicians, and women often rated complications as less severe than did men.

Just as patient perceptions of perioperative events are influenced by surgical indication and patient demographics,⁴ physician perceptions of surgical risk, benefit, and magnitude of consequences are likely to be related to surgeon training, indication for surgery, and patient selection. In this study, the aim was to explore the effect of surgeon training and indication for surgery (elective vs. nonelective) on perceptions of perioperative events unique to gynecologic surgery. By furthering our understanding of physician perceptions of complications, we can work toward improving preoperative counseling and expectation setting for patients. A standardized assessment tool for surgical complications may also allow surgeons to reliably compare postoperative outcomes, including patient satisfaction.

Materials and Methods

Following approval by the Loyola University Health System's Institutional Review Board, a task force of gynecologic surgeons who perform elective (benign gynecologists and urogynecologists) and nonelective procedures (gynecologic oncologists) was assembled to discuss potential intraoperative and perioperative complications as well as routine recovery events specific to gynecologic surgery. After discussion and review, 57 perioperative scenarios were compiled (Table 1): scenarios ranged from routine perioperative events, such as nausea and postoperative discomfort, to major operative complications, such as bowel perforation. The 57 scenarios were then grouped into 16 broader categories, listed in Table 2. The final list of scenarios was redistributed for review to a group of senior gynecologic surgeons to ensure construct validity.

Table 1.

Perceptions of Perioperative Scenarios

	% Complication			Coefficient of variation (Cv)
Scenario	All surgeons	Elective surgeons	Nonelective surgeons	All surgeons	Elective surgeons	Nonelective surgeons	Category of perioperative event
Change of access from laparoscopy to laparotomy	35.8	45.9	12.5	196	166	342	Change of surgical access
Change of access from hysteroscopy to laparoscopy	81.1	68.9	66.7	104	101	109
Change of access from hysteroscopy to laparotomy	88.7	89.2	87.5	47	46	54
Bladder injury, no change in access or prolonged catheterization	90.6	91.9	87.5	106	112	91	Bladder injury
Bladder injury, no change in access but prolonged catheterization	92.3	94.4	87.5	85	95	61
Bladder injury with change of access and prolonged catheterization	100	100	100	43	46	37
Unrecognized bladder injury resulting in vesicovaginal fistula	98.1	100	97.3	24	24	25
Ureteral injury, no change of access	93.9	94.1	93.3	65	65	64	Ureteral injury
Ureteral injury, change of access	100	100	100	37	39	33
Unrecognized ureteral injury resulting in ureterovaginal fistula	100	100	100	20	19	21
Bowel injury, no change in access or bowel resection	89.8	94.1	80	79	66	117	Bowel injury
Bowel injury no change in access but with bowel resection	98.0	97.1	100	44	41	47
Bowel injury with change of access and bowel resection	100	100	100	29	28	32
Bowel injury with change of access and bowel resection and ostomy	100	100	100	20	21	19
Vascular injury, no change of access	85.4	85.3	85.7	76	74	85	Vascular injury
Vascular injury requiring change of access	100	100	100	43	41	49
Intraoperative oral consultation from another service	2.1	0	7.1	0	0	379	Intraoperative consultation
Intraoperative surgical consultation from another service	8.5	6.1	14.3	378	500	253
Intraoperative blood transfusion	61.7	75.8	28.6	135	115	167	Blood transfusion
Postop blood transfusion	59.6	72.7	28.6	135	115	174
Transfusion reaction managed with Benadryl^®/Tylenol	63.8	72.7	42.9	126	111	152
Transfusion reaction requiring treatment beyond Benadryl/Tylenol	80.9	84.8	71.4	87	83	94
Postop nausea/vomiting resolved by discharge, hospital stay not affected	8.5	12.1	0	314	189	0	Postoperative nausea/vomiting
Postop nausea/vomiting resolved by discharge, but longer hospital stay	38.3	39.4	35.7	162	177	79
Postop nausea/vomiting treated as outpatient	23.4	24.2	21.4	228	223	248
Transient urinary retention resolved within 1 week postop	26.1	27.3	23.1	318	308	238	Urinary retention, urinary tract infection
Transient urinary retention for a few weeks requiring ISC	40.4	42.4	35.7	194	197	172
Postop UTI treated outpatient with antibiotics	48.9	51.5	42.9	155	163	133
Prolonged hospital stay for any reason	34.8	33.3	38.5	165	170	148	Prolonged hospital stay
Anticipated ICU transfer postop	10.6	12.1	21.4	268	275	247
Unanticipated ICU transfer postop	89.4	93.9	78.6	42	47	28
Death of patient	100	100	100	0	0	0
ER visit for any postop event	52.2	56.3	42.9	118	131	88	Readmission/reoperation
Hospital readmission for any postop event, no return to OR	93.5	96.9	85.7	60	64	53
Hospital readmission with return to OR	100	100	100	32	36	23
Interval reoperation/surgical treatment for any postop event	95.7	96.9	92.9	39	38	40
Suture/mesh erosion, managed as outpatient	95.7	96.9	92.9	67	74	49	Foreign body
Suture/mesh erosion, managed surgically	100	100	100	46	51	30
Postoperative pain	6.5	9.4	0	38	318	0	Common postoperative events
Postop constipation, managed as outpatient	4.3	6.3	0	600	500	0
Postop fatigue	0	0	0	0	0	0
Postop diarrhea, managed as outpatient	4.3	6.3	0	600	500	0
Postop abdominal distension, managed as outpatient	4.3	6.3	0	500	450	0
Poor cosmetic outcome of surgical wound, no revision	30.4	31.3	28.6	200	221	168
Sensory neuropathy, self-limited	47.8	53.1	42.9	165	165	152	Postoperative neuropathy
Sensory neuropathy, persistent	97.8	96.9	100	64	69	45
Motor neuropathy, self-limited	89.1	87.5	92.9	75	89	48
Motor neuropathy, resolved with conservative treatment (physical therapy)	97.8	100	100	44	49	32
Motor neuropathy, persistent	100	100	100	32	35	24
Postop ileus with prolonged hospitalization	47.8	56.3	35.7	144	131	175	Ileus/small bowel obstruction/abscess
Postop ileus/small bowel obstruction with NGT and prolonged hospitalization	84.8	96.9	57.1	62	64	54
Small bowel obstruction that required reoperation (did not resolve with NGT)	95.6	100	85.7	44	40	56
Pelvic abscess resolved with conservative mgmt.	93.5	100	78.6	61	55	78
Persistent SUI after anti-incontinence procedure	34.8	34.4	35.7	1.78	53	149	Lower urinary tract/urogynecologic symptoms
De novo SUI	43.5	40.6	50	143	153	121
Postop irriitative voiding symptoms	17.4	18.8	14.3	2.5	270	192
Postop UUI, self-limited	8.7	12.5	0	298	148	0
Postop UUI requiring treatment	52.2	50	57.1	120	133	101
Reoperation for persistent urinary retention	89.1	93.8	78.6	58	24	64

ISC, intermittent self-catheterization; UTI, urinary tract infection; ICU, intensive care unit; ER, emergency room; OR, operating room; NGT, nasogastric tube; SUI, stress urinary incontinence; UUI, urge urinary incontinence.

Table 2.

Intraclass Correlation Coefficients of Perioperative Events

	ICC Severity	ICC Clavien-Dindo
Change of access	0.261	0.269
	p<0.0001	p<0.0001
Bladder injury	0.122	0.174
	p<0.0001	p<0.0001
Ureteral injury	0.266	0.192
	p<0.0001	p=0.001
Bowel injury	0.274	0.269
	p<0.0001	p<0.0001
Vascular injury	0.578	0.357
	p<0.0001	p=0.001
Intraoperative consultation	0.191	0.154
	NS	NS
Blood transfusion	0.556	0.490
	p<0.0001	p<0.0001
Postoperative nausea/Vomiting	0.453	0.480
	p=0.002	p=0.001
Urinary retention & UTI	0.716	0.487
	p<0.0001	p<0.0001
Length/Service of hospital stay	0.014	0.015
	p=0.006	p=0.032
Readmission/reoperation	0.190	0.020
	p<0.0001	NS
Foreign body	0.550	0.034
	p<0.0001	NS
Common postoperative events	0.074	0.103
	p=0.026	p=0.005
Neuropathy	0.284	0.199
	p<0.0001	p<0.0001
Ileus, bowel obstruction, abscess	0.247	0.039
	p<0.0001	p=0.016
LUT/Urogynecologic symptoms	0.462	0.222
	p<0.0001	p=0.001

ICC, intraclass correlation coefficient; UTI, urinary tract infection; LUT, lower urinary tract.

Scenarios were electronically distributed to a sample of gynecologic surgeons from geographically diverse academic medical centers in the United States. Participating surgeons classified each scenario as either a complication or a routine perioperative event. For events classified as complications, surgeons rated the severity of the event on a numeric rating scale ranging from 1 to 100 and rated the event from I to V using the Clavien–Dindo scale.^3,4 Respondents also provided demographic information including age, gender, general and specialty board certification, monthly surgical volume, fellowship training (if applicable), and number of years in practice.

For analysis, surgeons were divided into two groups based on whether they performed elective gynecologic surgery (urogynecology, reproductive endocrinology and infertility, and general gynecology) or nonelective surgery (gynecologic oncology). PASW version 18 was used to calculate mean severity scores and standard deviations. In order to assess the degree of agreement among surgeons regarding a particular scenario, the coefficient of variability (Cv) for severity rankings of each scenario within and among surgeon groups was calculated. (The lower the Cv, the less spread there is in the distribution of severity scores and the higher the agreement.) In order to assess the reliability and reproducibility of the scenarios, the intraclass correlation coefficient (ICC) for each of the 16 categories was calculated. A statistically significant ICC reflects a high degree of reliability and reproducibility, indicating that the scenario was similarly interpreted by each of the respondents.

Results

Sixty-three of 82 (77%) contacted surgeons completed the web-based survey; surgeons performed a mean of 18 (range 3–90) cases per month and were on average 5 years (range 0–25) out of training. Sixty-four percent of respondents were board certified in obstetrics and gynecology. Twenty-nine percent (n=18) had completed fellowship training in female pelvic medicine and reconstructive surgery, 29% (n=18) had completed fellowship training in gynecologic oncology, and 11% (n=7) had completed fellowship training in reproductive endocrinology and infertility. The remaining 31% (n=20) were general gynecologists.

To ascertain whether each question was reliably and consistently interpreted by respondents, we calculated the intraclass correlation coefficient (ICC) for each of the 16 categories of scenarios (Table 2). The ICC was statistically significant (p<0.05) for all categories except “intraoperative verbal consultation from other surgical service,” indicating that there was high inter-rater reliability for all other questionnaire items. This was compared with the ICC values for the validated Clavien–Dindo scale, which revealed similar results.

Surgeon agreement regarding whether each specific perioperative scenario was a complication was then compared by calculating the Cv for each scenario. The range for Cv in th study was 0–600, with 0 indicating little variance or high agreement, and higher numbers indicating lower agreement (Table 1). Surgeons from all gynecologic training backgrounds expressed higher agreement on intraoperative events than on postoperative events (Cv range 0–96 vs. 32–318). There was also greater agreement about objective postoperative events with a measurable outcome, such as hospital admission (93.5% of surgeons considered this a complication, Cv 60) or repeatoperation (95.7% of surgeons considered this a complication, Cv 39), than about indolent patient-reported events, such as ileus (47.8% of surgeons considered this a complication, Cv 144), self-resolved sensory neuropathy (47.8% considered this a complication, Cv 165), or poor cosmetic outcome of surgical wound (30.4% considered this a complication, Cv 200).

For scenarios with high coefficients of variability (>100), the responses of nonelective (gynecologic oncologists) and elective (all other respondents) surgeons were compared, to find differences in perceptions of events based on the surgeon's training and clinical practice (Table 1). Nonelective surgeons classified conversion from laparoscopy to laparotomy, blood transfusion, and delayed return of bowel function as complications less frequently than did elective surgeons (36% vs. 12%, p=0.02; 62 vs. 29%, p=0.003; 85% vs. 57%, p=0.002), which likely reflect differences in surgical procedures and patient characteristics between these two groups. Whereas 100% of elective surgeons considered a pelvic abscess to be a complication, 21.4% of oncologists said it was not a complication. Oncologists were also more tolerant of postoperative pain, constipation and diarrhea, and abdominal distension than were elective surgeons (0 vs. 9.4%, 0 vs. 6.3%, 0 vs. 6.3%, 0 vs. 6.3%, p=0.026).

The responses of urogynecologists were analyzed separately from those of other gynecologic surgeons because of the specific nature of the complications encountered following quality of life surgery for pelvic floor disorders. Urogynecologists rated postoperative urinary symptoms (urinary retention, urinary tract infections (UTI), persistent or de novo stress incontinence, irritative symptoms, urgency incontinence) as more severe complications than did other surgeon groups (64–87% vs. 8–68%). All gynecologic surgeons, regardless of specialty, showed high agreement (96–100%) classifying postoperative events involving foreign body (mesh, suture) as complications (Cv=46, range 30–51).

Conclusions

This study found that surgeons' perceptions of perioperative events vary based on surgical indication and subspecialty training. Not surprisingly, surgeons who perform primarily quality-of-life elective procedures tend to emphasize and rate complications as more severe than do gynecologic oncologists performing nonelective cancer surgery. Most likely these differences in perception are reflected in different approaches to patient counseling and expectations, and may affect patient preparedness for and satisfaction following surgery.

Although all surgeons in this study, regardless of training or surgical indication, rated intraoperative complications as more severe than postoperative complications, patients often perceive postoperative onset of new symptoms more unfavorably than persistence of prior benign gynecologic problems or serious intraoperative events.² Less than 10% of surgeons rated pain, constipation, and diarrhea as complications. Conceivably, postoperative complications may be more bothersome to patients. For example, a patient may be aware of a vascular injury that is repaired intraoperatively only from the report of the surgeon, and experience little physical impact from it; however, a prolonged ileus often causes a patient considerable discomfort. However, current validated complications scales do not capture the magnitude and consequences of the postoperative course for patients.

Surprisingly, despite specialty-specific differences in perception of complications and their severity, nearly all surgeons identified foreign body or vaginal mesh exposure as complications. Permanent meshes bring new and different postoperative risks than those encountered with native tissue repairs; therefore, surgeons are obligated to identify and counsel patients about these unique risks and about strategies for managing mesh complications.^5,6 Patients often perceive the onset of new symptoms more unfavorably than the persistence of prolapse, and these findings suggest that gynecologic surgeons recognized the potential effects of mesh-related complications and their significance to patients.

A potential limitation of this study is the nonvalidated nature of the perioperative scenarios that comprised the questionnaire. In order to address this, the ICC was calculated as a measure of responder reliability or reproducibility for each category of scenarios. When the ICCs calculated based on the 1–100 numeric rating scale and the I–V Clavien–Dindo scale were compared, the same categories were statistically significant for each rating scale, indicating that the scenarios were reliably interpreted by respondents and that the responses were reproducible across the two complication rating systems. These data suggest that there are relevant differences in gynecologic ‘ perceptions of perioperative complications and their severity, which are most notable among gynecologic oncologists. Future studies assessing agreement between patients' and surgeons' perceptions of the severity of perioperative events may improve physician–patient communication and aid in preoperative counseling and alignment of presurgical expectations, and could be useful in creating a patient-oriented gynecologic surgery complication assessment tool.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Kenton

, Pham

, Mueller

, Brubaker

. Patient preparedness: An important predictor of surgical outcome. Am J Obstet Gynecol, 2007; 197:654.e1.

Pham

, Kenton

, Mueller

, Brubaker

. New pelvic symptoms are common after reconstructive pelvic surgery. Am J Obstet Gynecol, 2009; 200:88.e1.

Dindo

, Demartines

, Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg, 2004; 240:205.

Clavien

, Barkun

, de Oliveira

et al. The Clavien-Dindo classification of surgical complications: Five-year experience. Ann Surg, 2009; 250:187.

Nygaard

, Chai

, Cundiff

et al. Summary of research recommendations from the inaugural American Urogynecologic Society research summit. Female Pelvic Med Reconstr Surg, 2011; 17:4.

Haylen

, Freeman

, Swift

et al. An International Urogynecological Association (IUGA)/International Continence Society (ICS) joint terminology and classification of the complications related directly to the insertion of prostheses (meshes, implants, tapes) and grafts in female pelvic floor surgery. Int Urogynecol J, 2011; 22:3.