| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2001;72:515-520
© 2001 The Society of Thoracic Surgeons
a Infectious Disease Institute, Beer-Sheva, Israel
b Hospital Management, Beer-Sheva, Israel
c Quality Unit, Soroka University Medical Center and the Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel
Accepted for publication May 1, 2001.
Address reprint requests to Dr Borer, Department of Internal Medicine E, Soroka Medical Center, PO Box 151, Beer-Sheva 84101, Israel
e-mail: giladk{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. To identify contributors to increased deep-sternal infection rates in our institution, consecutive open-heart surgery patients were prospectively studied during two periods (75 and 40 days), including 66 and 40 patients, respectively. Active monitoring including 149 infection control practices was performed in the operating room and intensive care unit. End-points were deep-sternal infection rates and their relation to infection control practices.
Results. Mean age was 62 ± 11 years and 68% were males. Coronary bypass was performed in 82%. Clinical and surgical features were comparable, except that patients in period 2 were more likely to have heart failure (15% vs 1.5%, p = 0.01) and had a longer mean duration of surgery (277 vs 217 minutes, p < 0.005). Only 57 practices (38%) were adequately performed. The main categories showing inadequate practices were disinfection, traffic, hand-washing, and surgical attire of nonscrubbed personnel, anesthesiologists, and pump technicians. Many categories showed a statistically significant improvement between periods. Deep-sternal infection rates in prestudy and poststudy periods were 10% and 2.8%, respectively (p = 0.007).
Conclusions. Active monitoring among personnel involved in open-heart surgery resulted in a significant and sustained decrease in deep-sternal infection rates, through modification of human behavior and improvement of performance standards, probably mediated by the Hawthorne effect. Periodic active monitoring may be a valuable tool to achieve and even sustain such a decrease with tremendous implications on morbidity, costs, and quality of care.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
While SSIs in general have been reported to increase hospital stay and have a tremendous impact on the costs of care [6–8], DSIs are associated with even greater costs [9]. DSIs carry an enormous impact on patient morbidity [10], and it has been shown that infection is the most expensive complication after coronary bypass surgery and that the cost of care for patients with DSI rises threefold [11].
Possible host, surgical, and nosocomial risk factors associated with postoperative infection in cardiac surgery have been extensively studied, mostly in a retrospective fashion. However, the association between compliance of hospital personnel with infection control practices (ICP) and the rates of infection after OHS, especially DSI, has not been given much attention, and the relative contribution of each of the above-mentioned factors to the development of DSI remains unknown.
During the last 2 years, a relatively high rate of DSI (approximately 10%) was detected among OHS patients in our institution. We thus undertook a prospective surveillance study in order to evaluate potential contributors to the development of DSI in our patients. Special emphasis was placed on ICP of personnel, using a unique active monitoring approach, with the intention of performing a subsequent intervention for the correction of such risk factors.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
We prospectively followed consecutive adult patients (over 18 years of age) undergoing OHS in our institution between September 23, 1998 and February 5, 1999. Relevant clinical and surgical data as well as outcome measures were collected (see below). Active monitoring of ICP was performed on a random sample of patients during the surgical procedure and throughout their stay in the surgical intensive care unit.
Patient characteristics
The following were collected for each of the study patients: age; gender; duration of hospitalization before surgery, in the surgical intensive care unit (SICU), and surgical ward after surgery; prior admission to an intensive coronary care unit (ICCU); and underlying medical conditions including smoking, diabetes, obesity, chronic obstructive pulmonary disease (COPD), peripheral vascular disease (PVD), chronic renal failure (CRF), congestive heart failure (CHF), diagnosis of active infection before surgery, and the American Society for Anesthesiology (ASA) score.
Surgical characteristics
The following data were collected: type of procedure (coronary bypass, valve replacement, or both), time of session (morning or afternoon), type and number of grafts used (saphenous vein graft [SVG] or internal mammary artery [IMA]), timing of surgery (emergency or elective), duration of procedure (skin-to-skin), as well as cardiopulmonary bypass time and aortic-clamp time, and the need for intraaortic balloon counterpulsation pump (IABP) insertion. In addition, the type, timing, and dosing of all antibiotic prophylaxis doses given were recorded.
Outcome
We focused on DSI as the study end-point because of its impact on patient morbidity and cost of care. Organ/Space SSIs were included in the DSI group, and leg SSIs were not studied. As this hospital is the sole referral facility in the region, no cases of wound infection could have gone unnoticed. We assessed the DSI rate during the study, as well as during the 6 months after its completion. These rates were compared with preexisting surveillance data obtained during an 18-month period before the initiation of the current study. These data included wound infection surveillance for 563 operated patients obtained in a method similar to the present study, excluding active monitoring. Deep-sternal and Organ/Space SSIs were diagnosed according to the definitions of the Centers for Disease Control and Prevention (CDC) [12]. Surveillance of DSI cases was performed using direct observation by surgeons and infectious disease consultants and by indirect methods such as laboratory and patients’ records and readmissions due to infection [13]. Surveillance was continued regularly for an additional 6 months after study completion.
Active monitoring of ICP
Active monitoring was performed in both the operating room (OR) and SICU for randomly selected patients, by 3 trained full-time registered OR nurses, highly experienced in the care of OHS patients, using a specially developed questionnaire. OR monitoring was performed continuously throughout all activities related to the same procedure. SICU monitoring was performed in several distinct time periods and focused on activities related to the same patient whose operation was monitored earlier. Personnel were not aware of the content of questionnaires, and no feedback was given during or after each monitoring session. Moreover, personnel were not notified in advance of which sessions were about to be monitored. No educational intervention whatsoever was performed during the study period.
Overall, the questionnaires included 149 ICP (available from the authors) related to surgical, nursing, bypass pump, and anesthesia personnel as well as orderlies and cleaning workers, covering a wide spectrum of infection control issues. Desired performance of personnel was based in most instances upon the ASA and CDC recommendations [13, 14]. Desired performance for items not included in these guidelines was based on the local recommendations issued by the Israeli Ministry of Health. All ICP represented yes/no questions aimed at recording actual activities with minimal or no subjectivity and observer-related bias.
Briefly, monitoring of OR activities included 121 ICP in six categories, including: (A) environmental disinfection (22 items); (B) ICP of anesthesiologists and cardiopulmonary bypass pump (CBP) technicians (21 items); (C) surgical attire (24 items); (D) Surgical scrubbing (24 items); (E) Sterile preparation and draping (17 items); and (F) ICP of scrubbed personnel, surgeons, and OR traffic (14 items). Monitoring in the SICU included 28 ICP relating to: (A) environmental disinfection (10 items), and (B) ICP of SICU personnel (18 items).
Owing to administrative reasons, active monitoring could not be performed between December 14, 1998 and December 26, 1998. As a result, active monitoring took place during two time periods: period I, September 23, 1998 to December 13, 1998 (75 days); and period II, December 27, 1998 to February 5, 1999 (40 days).
Statistical analysis
Data were analyzed using the Epi-Info software, version 6.02 (CDC, Atlanta, GA). The clinical and surgical characteristics of all patients throughout the study period as well as outcome measures were analyzed. As monitoring took place during two designated time periods, these were compared using the 
2, Mantel-Haenszel, two-tailed analysis of variance, and Fisher’s exact tests, as appropriate for both clinical and ICP data in order to clinically correlate these parameters. To eliminate any bias, patients operated between periods were excluded from the latter analysis. The characteristics of patients suffering from DSI were compared with those not infected.
Compliance rates observed during active monitoring were compared between periods using the Mantel-Haenszel and 2 tests, as appropriate. We assumed that the expected performance standard should be 100% compliance, as each of the items may possibly increase the risk for SSI. We arbitrarily defined significant noncompliance as a rate less than 90%. For this purpose, ICP were grouped into clusters, each representing an infection-control issue. Compliance was calculated by dividing all observations related to adequately performed ICP in each category, by the sum of monitored observations related to these ICP.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
The overall mean age of patients was 62 ± 11.5 years (median 63, range 30 to 82 years), and 80 (68%) were men. The most prevalent host risk factors were a mean hospital stay of 7.8 ± 7.8 days before surgery, prior admission to a coronary care unit (21%), diabetes (22%), smoking (32%), and obesity (19%). Twelve patients (10%) were diagnosed with an infection before surgery. The clinical features of patients in both time periods were comparable, except that patients in the second period were more likely to suffer from congestive heart failure (15% vs 1.5%, p = 0.01)
Surgical characteristics
Most patients underwent coronary bypass surgery (81%) involving SVG + IMA grafts (90%). Operations were mainly elective (90%), and only 4% of patients were assigned an ASA score of IV. The mean duration of surgery (skin-to-skin) was 242 ± 89 minutes (median 240, range 9 to 525 minutes) and the mean bypass time was 106 ± 37 minutes (median 105, range 42 to 295 minutes). All patients received intravenous antibiotic prophylaxis, cefuroxime 750 mg in 97% and vancomycin 1 g in 3%. The first dose was given at a mean of 56 ± 33 minutes before the surgical incision (median 57, range 5 to 220 minutes) and the second 242 ± 124 minutes after the first (median 220, range 50 to 690 minutes). The mean number of doses was 6.2 ± 2.6 per patient, with a mean of 2.4 ± 1.0 antibiotic days. There were no significant differences between patients in both periods, except that patients operated during period II had a significantly longer mean duration of surgery (277 ± 96 vs 217 ± 74 minutes, p = 0.0024).
Outcome
The mean hospital stay was 18.3 ± 10.5 days overall (median 16, range 8 to 63 days) and 12.7 ± 7.1 days (median 10, range 6 to 63 days) in the cardiothoracic surgical ward, including SICU stay. The overall rate of DSI alone, not including other infectious diagnoses, was 5.1%, a decrease from 7.6% to 0% between period I and period II (p = 0.14). DSI cases were evenly distributed throughout period I. No DSI cases occurred during the 2 weeks free of monitoring. The rate of nosocomial infection other than SSI was 6.8%, (3% in period I and 15% in period II, p = 0.051). The in-hospital mortality rate from all causes was 7.5%. There were no differences in terms of hospital stay or mortality between the two groups.
According to our surveillance data, 56 cases of DSI occurred in 563 patients operated upon during the 18-month period preceding the initiation of the current study, while during a period of 6 months after the completion of the study, four DSI cases were detected out of 142 OHS procedures. Thus, the pre- and poststudy DSI rates were 10% and 2.8%, respectively (p = 0.007).
Risk factors for DSI
When patients with DSI were compared with those without DSI, none of the host risk factors was found to be significantly associated with DSI, except for a trend towards increased risk with obesity (50% vs 17%, respectively, p = 0.077) and an infection before surgery (33% vs 9%, p = 0.09). Neither any of the surgical characteristics nor antibiotic prophylaxis were found to be significantly associated with an increased risk for DSI. Patients with DSI had a significantly longer mean total hospital stay (27 ± 12 vs 18 ± 10 days, p = 0.035), most of which could be attributed to prolonged stay in the cardiothoracic surgical ward (16 ± 7 vs 7 ± 5.5 days, p < 0.001). DSI was not associated with an increased mortality rate.
Active monitoring of ICP
Active monitoring in the OR and SICU was performed for 57 randomly selected patients, 40 in period I and 17 in period II. Data regarding the 149 ICP were available for all monitored patients. Of these, only 57 (38%) were performed adequately in 100% of observations (maximal compliance). Compliance rates for the various ICP categories in both study periods are presented in Table 1.
|
A statistically significant improvement in the overall compliance rate of personnel with ICP was noted between the two time periods, for both OR and SICU activities overall, as well as for ICP recommended and not recommended by guidelines, and those related to environment and disinfection in both the OR and SICU. An improvement, though not statistically significant, was noted for anesthesiologists, CBP technicians, and OR traffic. There were no significant declines in personnel compliance.
|
Comment |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
The exact mechanism by which DSI develops is unknown and involves multiple factors [15, 16], the relative impact of which may differ between institutions, and the weight and importance of each individual factor are difficult to assess. Many studies have established risk factors for DSI. Most were retrospective, seldom using a case-control design, and largely influenced by differences in patient populations, geographic location, methodology, definitions, and end-points [4, 15, 17, 18].
None of the major host risk factors was found significantly related to DSI in our study, possibly due to the small number of infections. Obesity and an infection before surgery, however, demonstrated a trend towards increased DSI risk, as has been shown previously [5, 15, 17, 18]. Of the surgical features studied, a longer mean duration of surgery, a clear risk factor for DSI, was noted during the second period despite the decrease in DSI. Such duration, however, although not optimal, still lies within acceptable percentiles according to the NNIS risk index [19]. Additionally, antibiotic prophylaxis was infused a median of 60 minutes before surgery during both periods. Although it has been shown that prophylaxis given within 2 hours before incision reduces SSI rate [20], current recommendations favor the optimal infusion of cephalosporins 30 minutes before incision [13]. Although no DSI cases occurred during period II, these findings merit correction.
It has been shown that specific infection control interventions may reduce SSI rate by 20% to 50% [6, 21–23] and that behavior of operating theater personnel regarding decontamination, hand antisepsis, and compliance with universal precautions are major risk factors for SSI [24–26]. Moreover, a recent survey of operating theater staff revealed an alarmingly limited knowledge about ICP [27], and implementation of an infection control program involving 10 ICP issues significantly decreased the overall SSI rate (but not DSI) after OHS from 12.4% to 8.9% [10]. Thus, ICP of personnel have a major impact on SSI.
We have discovered many inadequate ICP and poor compliance rates in numerous categories, especially by anesthesiologists, CBP technicians, nurses, and orderlies. Significant noncompliance rates were also found in categories such as handwashing, OR traffic, and disinfection, as well as both OR-related ICP addressed and not addressed in guidelines. These findings emphasize the importance of ICP performed in the "outer circle," away from the sterile surgical field, by nonsurgical team members, and that at least some categories, for which there is currently no official recommendation, are potential contributors to DSI.
Awareness of being studied or observed may alter the way in which a person acts or performs, through intentional or unintentional cues, and is further enhanced by demand for change in performance or a feedback mechanism. This phenomenon has been termed the "Hawthorne effect" [28, 29]. The effect derives its name from studies performed from 1927 to 1932 in the Hawthorne Works Plant of the Western Electric Company in Chicago that have shown that observation alone resulted in increased efficiency of workers despite many work condition changes. The effect was sustained even after return to prestudy conditions [30]. We believe that our results can be explained by the presence of such an effect that might in fact translate into an actual improvement of clinical end-points, namely DSI.
It has been claimed that no single intervention has consistently been shown to increase or sustain improved compliance with respect to ICP [31]. If our results will be validated in future studies, active monitoring of ICP may prove a useful tool for both detecting inadequate ICP and influencing the incidence of DSI after OHS. As human nature has been deemed an important cause for failure to implement ICP [32], periodic active monitoring may serve as a reminder, sustain the effect of intervention, and provide feedback. This should be performed carefully, as imposition of ICP policies may result in a negative effect [32]. In our institution, more than 6 months after completion of the study and without continued monitoring, the significant reduction in DSI rates is still sustained.
DSIs involve substantial morbidity related to prolonged hospitalization and a risk for nosocomial infection, drug adverse effects, and repeated operations. The costs of care are also greatly affected, as each DSI episode leads to an increase of $23,000 in health expense [33]. On the contrary, the cost of periodic monitoring is surprisingly low (requiring the assignment of two nurses for a whole week, every 3 months), and may thus prove cost-effective even if only one case of DSI is prevented annually.
Some potential limitations of this study are worth mentioning. This study involved a relatively small patient sample, which might influence statistical power, especially for calculating risk factors for DSI unrelated to ICP (as the latter did involve a large number of observations) and outcome. Moreover, establishing a clear cause-and-effect relationship between ICP improvement and DSI rates requires a larger sample of observations, preferably by multivariate analysis. Nevertheless, the association demonstrated between ICP improvement and sustained decrease in DSI argues that our findings did not occur by chance and thus warrant further study.
We exclusively studied DSI and our results cannot be extrapolated to complications such as superficial sternal or leg SSI. Moreover, an unexpected trend toward increased non-SSI nosocomial infections was noted during period II. This may be attributed to increased rates of ventilator-associated pneumonia among patients with severe heart failure operated upon during that period.
In conclusion, active monitoring of infection control practices among personnel involved in OHS resulted in a significant and sustained decrease in DSI rates, through modification of human behavior and improvement of performance standards, probably mediated by the Hawthorne effect. Periodic active monitoring may prove a simple, useful, and cost-effective tool in this setting. Further study is needed in order to validate the role of active monitoring as an essential part of procedures whose infective complications may cause devastating consequences.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. A. Ferraris, F. H. Edwards, D. M. Shahian, and S. P. Ferraris Risk Stratification and Comorbidity , January 1, 2008; 3(2008): 199 - 246. [Full Text]
|
|
|
|
|
|
V. A. Ferraris and S. P. Ferraris Risk Stratification and Comorbidity , January 1, 2003; 2(2003): 187 - 224. [Full Text]
|
|
|
|
 |
|
 T. M. Fleck, M. Fleck, R. Moidl, M. Czerny, R. Koller, P. Giovanoli, M. J. Hiesmayer, D. Zimpfer, E. Wolner, and M. Grabenwoger The vacuum-assisted closure system for the treatment of deep sternal wound infections after cardiac surgery Ann. Thorac. Surg., November 1, 2002; 74(5): 1596 - 1600. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
