Stakeholder Engagement Significantly Decreased Colorectal Surgical Site Infections

Abstract

Background:

Patients and care providers raised concerns about the increased incidence of colorectal surgical site infection (SSI) at a community hospital in Baltimore compared with peer institutions.

Patients and Methods:

A preliminary analysis was performed that identified several modifiable targets for interventions to reduce SSIs in this patient population. The intervention focused on wide engagement of all stakeholder groups across the spectrum of care including physicians, pharmacists, nurses, administrators, and patients. The engagement process involved hospital-wide educational sessions, adoption and implementation of the best clinical guidelines, and utilization of the electronic medical record system to reinforce compliance and ensure quality control. Data for SSIs in colorectal surgical procedures were collected prior to the intervention (January 1, 2017 to March 31, 2018) and after implementation (April 1, 2018 to October 31, 2018).

Results:

A total of 355 cases (229 pre-intervention group, 126 post-intervention group) met the inclusion criteria; the two groups were comparable with respect to all the key parameters except the procedure type and use of endoscopy. Multivariable logistic regression modeling was utilized to evaluate the effects of the stakeholder engagement intervention while adjusting for potential confounders. The incidence of colorectal SSIs was substantially lower after the intervention (2.78% vs. 8.73%, p = 0.02). This reduction was robust to adjustment for covariates in regression modeling (p = 0.04).

Conclusions:

Informed stakeholder engagement helped bring cohesion to the inherently fragmented elements of the care delivery model and was associated with decreased incidence of colorectal SSIs.

Surgical site infections (SSIs) continue to pose a multifold challenge because of many host and treatment factors that have been reviewed [1]. The literature abounds with several thousand studies that identified risk factors and reported on quality control initiatives that aimed to address this elusive problem [2]. This complication is likely to loom larger in the future as the healthcare landscape is redefined by more complex patients, multiple comorbidities, a projected increase in surgical procedures, and an expected increase in the emergence of antimicrobial-resistant pathogens. In light of this, the human [3] and economic [4] burden associated with SSIs, coupled with the relatively modest progress in fighting them, necessitate a more efficacious approach, especially in colorectal surgical procedures in which SSI rates surged up to 25%, the highest among other surgical procedures [5, 6].

The latest systematic review and guidelines by the U.S. Centers for Disease Control and Prevention (CDC), inclusive of 5,700 articles, underscored the paucity of robust evidence for several clinical practices [7]. These gaps in evidence were addressed by a panel of experts in the field [8] and their findings highlighted two important observations: the complex multifactorial nature of SSIs and more importantly, the inconsistent application of all SSI reducing measures [1]. The concept of surgical care bundles in colorectal surgery was introduced and showed reduction of SSIs compared with standard treatments, yet these bundles are neither standardized nor applied widely. The multidisciplinary approach to mitigate SSIs has slowly evolved but the quality of evidence and elaboration on the details of this approach are currently limited [9].

Despite several advances in the understanding and prevention of SSI, interventions appear to have reached the point of diminishing returns. One of the limitations on further improvement in SSI prevention is lack of stakeholder engagement [10]. This article reports on applying the concept of stakeholder engagement at a 500-bed, non-profit, teaching, inner-city community hospital to reduce colorectal deep and organ/space infection. The initiative was started after the 2017 National Healthcare Safety Network (NHSN) report indicated Sinai Hospital's poor performance compared with peer institutions, with respect to deep and organ/space infection in patients with defined colorectal procedures. A total of 193 procedures were performed with 16 procedures associated with SSIs, exceeding the predicted 7.3 cases, resulting in a standardized infection ratio (SIR) of 2.19; (95% confidence interval [CI], 1.30, 3.19) [11]. The concept of informed stakeholder engagement, was promulgated initially by the International Finance Corporation, an agency of the World Bank Group, to improve social performance [12]. The concept was implemented in parallel within a community hospital setting to decrease SSIs. All institutional stakeholders from direct care providers to the senior leadership were identified and actively involved to mitigate this problem with the underlying presumption that the healthcare institution possessed the resources but lacked the effective framework to reach this goal. In this context, stakeholder engagement was an endeavor to promote education, communication, and shared responsibility.

Patients and Methods

Data on SSIs in colon surgery were downloaded from the NHSN, which provided the hospital's SIR and comparative data for the state of Maryland. The SIR was calculated by dividing the number of observed infections by the number of predicted (expected) infections. The number of predicted infections was calculated using SSI probabilities estimated from multivariable logistic regression models constructed from NHSN data. Inpatients who were ≥18 years of age who underwent colorectal procedures and presented with deep incisional primary SSIs and organ/space SSIs within 30 days of the procedure on re-admission and post-discharge surveillance were included and the logistic regression model adjusted for diabetes, American Society of Anesthesiologists (ASA) score, gender, age, body mass index (BMI), oncology hospital, closure techniques, and use of endoscopy [7].

Our framework to implement stakeholder engagement, as advocated by the Agency for Healthcare Research and Quality [13], and proposed by others [14], aimed to raise awareness, build collaborations, and maintain sustainability. In relation to our model, awareness was raised by the senior leadership and the chairmen of each department involved. Collaboration was embodied by the taskforce that was charged with identifying solutions and gathering momentum for their implementation. Finally, sustainability was ensured by holding monthly meetings of the taskforce to track progress. The details of each of these elements in relation to our stakeholder engagement model are described below.

After deliberation within the hospital leadership, a taskforce was deployed to assess the contributing factors, identify and engage all stakeholder groups, and track progress on a quarterly basis. The taskforce included senior leaders, surgeons, perioperative nursing staff, infection control specialists, pharmacists, central sterile and environmental services members, educators, and quality/risk personnel. Several institution-wide educational sessions and lectures were held to raise awareness about SSIs and call for action by every provider within their domain of expertise. The best clinical practices, in accordance with CDC guidelines, were implemented in collaboration with the taskforce members. The guidelines included modifying intra-operative process by using incision protectors, using a separate closure set, minimizing open surgical approaches and encouraging laparoscopic/robotic-assisted procedures, and decreasing the operative time, limiting operative hypervolemia, administering bowel preparation with pre-operative oral and intra-operative systemic antibiotic agents per guidelines, using chlorohexidine wipes, and maintaining normothermia.

The next step was to develop a communication strategy and ensure buy-in from all the stakeholders of all the guidelines and best practices that were adopted. Interdisciplinary communication and shared responsibility were highly encouraged throughout the process. The strategy focused on securing buy-in by all stakeholders including surgical team members, anesthesia team members, nursing staff, administrators, and patients to implement these guidelines. The departments of surgery, anesthesiology, and nursing services took the initiative to educate their respective staff on the best clinical practices, encourage communication, and emphasize responsibility. Educational sessions were held to inform surgery attending and resident physicians on proper classification and documentation of surgical incisions, patient comorbidities, and whether infection was present on admission. Similarly, the anesthesia team members highlighted adherence to maintaining normothermia, re-dosing of antibiotic agents, and glycemic control. Charge nurses in the operating suites and the hospital units educated their respective staff on incision evaluation, incision care, and communicating incision-related issues to the physician providers. The educational sessions began in January 2017 and continued through March 2018. Concurrently, a new electronic medical record order set was devised and implemented to help control perioperative glycemia and the anesthesia paper record was also fully migrated to the electronic platform. The institution experienced a surge of new hires during the same time period and educational goals were reiterated to avoid gaps in communication. The intervention was maximized by March 2018, as manifested by completion of all the educational sessions and assignments to each member of the taskforce. Last, the taskforce met on a regular basis and tracked progress. Suboptimal performance and need for support were addressed by the senior leadership as needed.

Data Collection and Statistical Analysis

Patient data were downloaded from the Maryland Health Care Commission (MHCC) database from January 2017 through March 2018 (pre-intervention), and April 2018 through October 2018 (post-intervention). A total of 368 records of colon surgery cases (per the definition set by the Centers for Disease Control and Prevention) were retrieved from search engine TheraDoc in January 2019. Patients younger than 18 years of age were excluded from the analysis. Patients in the analysis were characterized with respect to age, gender, procedure type (open, laparoscopic, robotic), ASA classification (1–4), wound classification (2–4), overall risk index, use of intra-operative endoscopy, acuity (emergent versus elective), type of closure (primary vs. delayed), presence of diabetes mellitus, BMI, smoking status, use of antibiotic agents, socioeconomic status (insurance plan as a surrogate measure), length of the surgical procedure, intra-operative hypothermia, surgeon type (specialist vs. generalist), and the incidence of SSI. The normality of continuous variables was assessed via the Kolmogorov-Smirnov test. Surgical site infection incidence and patient characteristics were compared pre- and post-intervention using two-tailed t-tests and χ² tests or Fisher exact test for continuous and categorical variables, respectively.

Multivariable logistic regression adjusting for the covariates listed above was performed. Statistical significance was defined as p ≤ 0.05. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC).

Results

Three hundred fifty-five patients (229 pre-intervention group, 126 post-intervention group) were included in the analysis. Table 1 characterizes and compares the pre- and post-intervention study samples on demographic, surgical characteristics, and risk factors. In brief, the pre- and post-intervention samples were similar across all covariates with the exception of procedure type (open vs. laparoscopic vs. robotic assisted) and use of intra-operative endoscopy (colonoscopy), intra-operative hypothermia, and socioeconomic status. This decrease in SSI incidence after the intervention was robust to adjustment for all covariates in logistic regression modeling (intervention maximum likelihood estimate = −0.06, p = 0.04).

Table 1.

Demographic Characteristics of the Patients in the Pre-Intervention Group (n = 229) and Post-Intervention Group (n = 126)

Parameter	Parameter	Pre-intervention	Post-intervention	p
		n = 229	n = 126
SSI incidence		8.73	2.78	0.02
Age (y)	Mean ± SD	64.5 ± 15.45	63.93 ± 16.30	0.7
Duration (min)	Mean ± SD	280 ± 171	311 ± 195	0.11
Gender	Male	43%	40%	0.6
	Female	57%	60%
Procedure type	Open	63%	50%	0.002
	Laparoscopic	18%	35%
	Robotic	19%	15%
ASA classification	1	1%	2%	0.2
	2	32%	39%
	3	54%	43%
	4	11%	17%
	5	2%	0%
Wound classification	2	60%	57%	0.7
	3	24%	24%
	4	16%	19%
Risk index	1	8%	3%	0.2
	2	33%	40%
	3	44%	44%
	4	14%	13%
Use of endoscopy	Yes	42%	54%	0.04
Emergent procedure	Yes	13%	17%	0.3
Primary closure	Yes	93%	95%	0.5
Diabetes mellitus	Yes	26%	33%	0.2
BMI >30	Yes	34%	41%	0.2
Caucasian race	Yes	46%	46%	0.97
Smoker status	Yes	25%	24%	0.86
Antibiotic use	Yes	90%	92%	0.58
Insurance (Medicare?)	Yes	51%	55%	0.51
Specialist	Yes	59%	67%	0.15
Hypothermia	Yes	10.6%	19.8%	0.02

p values were calculated using unpaired t-tests (continuous parameters) and χ² or χ² tests (categorical parameters).

SSI = surgical site infections; SD = standard deviation; ASA = American Society of Anesthesiologists; BMI = body mass index.

Table 2.

Regression Analysis for the Parameters Analyzed in the Study

Parameter	Estimate	Standard error	p
Intercept	0.1054	0.1445	0.4656
Post-intervention	−0.0573	0.0274	0.0362
Pre-intervention	Referent
Age	-0.003	0.0094	0.7461
Male gender	−0.0736	0.0264	0.0053
Female gender	Referent
Diabetes mellitus (yes)	0.0247	0.0308	0.4224
Diabetic (no)	Referent
BMI >30	0.0389	0.028	0.1646
BMI <30	Referent
Primary closure (no)	0.0168	0.0582	0.7725
Primary closure (Yes)	Referent
ASA category 3	0.0607	0.1276	0.6344
ASA Category 2	0.0355	0.1267	0.7793
ASA Category 1	Referent
Procedure (robotic)	−0.0083	0.0518	0.8724
Procedure (laparoscopic)	−0.0079	0.0458	0.8628
Procedure (open)	Referent
Endoscope (yes)	−0.0425	0.0433	0.3265
Endoscope (no)	Referent
Caucasian (no)	0.0112	0.0268	0.6775
Caucasian (yes)	Referent
Smoker (no)	−0.043	0.0317	0.1756
Smoker (yes)	Referent
Antibiotic use (no)	0.0846	0.0448	0.0592
Antibiotic use (yes)	Referent
Hypothermia (no)	0.0051	0.0373	0.8904
Hypothermia (yes)	Referent
Specialist (no)	−0.0416	0.0291	0.1521
Specialist (yes)	Referent

P values were calculated using unpaired t-tests (continuous variables) and χ² or χ² tests (categorical variables). The scale parameter was estimated by maximum likelihood.

BMI = body mass index; ASA = American Society of Anesthesiologists.

Discussion

Several insights were gleaned from this stakeholder engagement intervention aimed at reducing the relatively high incidence of SSIs at a community hospital. This intervention reduced the infection rate substantially, even after adjustment for covariates and possible confounders. To reiterate, no novel clinical practice/measure was introduced, but all the existing evidence-based measures shown to work in a multitude of previous high-quality studies and controlled clinical trials [1,15] were adopted within the framework for this intervention. This suggests that in some instances, formal engagement with stakeholders in implementing best practices is sufficient to improve critical outcomes such as SSI.

This intervention was an experiment in collaboratively modifying the organizational behavior and examining its impact on a relatively poor outcome, in comparison to peer institutions. Toward that goal, the aims were informing and engaging all stakeholders, creating the right milieu, eliminating cultural barriers, and cohesively promoting the use of best clinical practices. The concept of stakeholder engagement has been adopted in several business and management models to improve social performance [12] and is slowly but steadily emerging in the healthcare field [16]. The stakeholder engagement model appears well suited to translational quality outcomes research endeavors in real-world practice settings [17] and has been promoted by the National Institutes of Health (NIH) [18] and the Patient Centered Outcomes Research Institute (PCORI) [19]. The utility of this concept was also illustrated in several studies [20 –22]. However, to date, applying this concept to decrease SSIs has not been reported and this article is the first to show the utility of stakeholder engagement to achieve clinical improvement in SSIs.

In the attempt to apply this model, stakeholders were envisioned to include all medical and administrative staff whose actions could modulate the outcome, driven by different incentives. For example, the public reporting of SSIs on a state and national level was an incentive for the senior administrative leadership to mitigate this problem considering the relatively high rates of SSI at this institution prior to implementing the intervention. This goal was also aligned with the interests of other stakeholders, such as surgeons and anesthesiologists, who are affected by public reporting. The reimbursement structure was another identified driver. The educational seminars sought to engage all participants by encouraging them to voice their ideas and take ownership of what they could contribute within their respective role(s). For example, anesthesiologists identified proper dosing of antibiotics, maintenance of normothermia, and glycemic control. Similarly, surgical residents ensured adherence to the preoperative oral bowel regimen and antibiotics. Likewise, operating room staff ensured compliance with using incision protectors and provision of a separate closure table.

Admittedly, achieving buy-in from all stakeholders was not easy and required a top-down leadership approach to enforce in certain cases. The creation and implementation of the perioperative glucose protocol was a case in point. The initial attempts to integrate this protocol were considerably resisted by the anesthesia providers because it required additional electronic documentation, which interfered with an already laborious paper record documentation and pre-existing workflow. However, engaging the senior leadership and aligning this goal with the institution-wide goal of paperless documentation helped eventually eliminate this barrier.

Although the findings of this stakeholder engagement intervention were encouraging, several key limitations should be acknowledged. First, several factors unrelated to the stakeholder engagement intervention might have influenced the reduction in SSI rates after the intervention. For instance, there was an increase in the use of the laparoscopic compared with the open approach and the use of endoscopy during the post-intervention period. It is possible that these changes may have partially explained the decrease in SSI rate, because both of these factors have been associated with decreased SSIs in previous studies [15,23]. However, the reduction in SSI rates was robust to adjustment for this factor and other key covariates in logistic regression modeling. Accordingly, they are not likely to have confounded the associations reported. Another limitation is the relatively small sample size of the first 126 cases reported after implementation of the intervention. However, this sample size in addition to the 229 cases immediately preceding the intervention as a basis of comparison, was sufficient for the multivariable regression modeling that was used to demonstrate the robustness of the findings. Several other factors such as the socioeconomic status, smoking status, diabetes, intra-operative hypothermia, and proper use of antibiotic agents (in compliance with Surgical Care Improvement Project recommendations) were tracked and found to have no effect on the final outcome in the regression model. Admittedly, several other factors such as the use of ostomy, changes in practice, and operating surgeons could explain the observed trend in this study, none of these factors were markedly different prior to and following the adoption of new guidelines. By process of elimination, we were unable to identify a specific factor that could explain this finding, further strengthening our hypothesis that stakeholder engagement was the driving factor in this study. The multivariable regression analysis failed to identify any of the factors tracked in this study as a driver of the difference in outcomes. Even though our surrogate measure for the socioeconomic status was suboptimal, it failed to eliminate the difference in the outcome of this study after it was included in the regression analysis.

Future directions of this work include attempting to apply this concept to other quality improvement initiatives at this institution and create a loop feedback mechanism to evaluate this framework and the interventions further. Pertinent to this goal is longitudinal tracking of patients and monitoring of compliance. An attractive opportunity would be to increase patient engagement in the next phase of this process.

In conclusion, this is the first article to report on the success of incorporating the principles of informed stakeholder engagement to significantly lower the incidence of colorectal SSIs. Completely teasing apart the contribution of modulated specific SSI etiologic factor(s) versus that of stakeholder engagement may not be feasible. However, it is imperative to recognize that the SSI factors (building blocks) and the stakeholder engagement framework (scaffolding) are inseparable constituents of the body (building) of this approach. The application of this framework resulted in substantial reduction in the SSI rate at this hospital and improved the quality of care delivered to its patients, suggesting that this approach may be worthy of consideration at other institutions.

Footnotes

Funding Information

Funding was provided by the Department of General Surgery at Sinai Hospital of Baltimore. No external funding or grants were received.

Author Disclosure Statement

The authors have no conflicts of interest or financial ties to disclose.

References

Fry

. The prevention of surgical site infection in elective colon surgery. Scientifica (Cairo), 2013; 2013:896297.

Fry

. Fifty ways to cause surgical site infections. Surg Infect, 2011; 12:497–500.

Stone

, Braccia

, Larson

. Systematic review of economic analyses of health care-associated infections. Am J Infect Control, 2005; 33:501–509.

Anderson

, Kirkland

, Kaye

, et al. Underresourced hospital infection control and prevention programs: Penny wise, pound foolish?. Infect Control Hosp Epidemiol, 2007; 28:767–773.

Hedrick

, Sawyer

, Hennessy

, et al. Can we define surgical site infection accurately in colorectal surgery?. Surg Infect, 2014; 15:372–376.

, Ko

, Hall

, et al. A comparison of 2 surgical site infection monitoring systems. JAMA Surg, 2015; 150:51–57.

Berrios-Torres

, Umscheid

, Bratzler

, et al. Centers for Disease Control and Prevention Guideline for the Prevention of Surgical Site Infection, 2017. JAMA Surg, 2017; 152:784–791.

Itani

KMF

, Dellinger

, Mazuski

, et al. Surgical site infection research opportunities. Surg Infect, 2017; 18:401–408.

Gillespie

, Kang

, Roberts

, et al. Reducing the risk of surgical site infection using a multidisciplinary approach: An integrative review. J Multidiscip Healthc, 2015; 8:473–487.

10.

Stelfox

, Niven

, Clement

, et al. Stakeholder engagement to identify priorities for improving the quality and value of critical care. PLoS One, 2015; 10:e0140141.

11.

Reports MHCQ. Surgical Site Infections. 2017. https://healthcarequality.mhcc.maryland.gov/public/ssi (Last accessed July 13, 2020).

12.

International Finance Corporation. Stakeholder Engagement: A Good Practice Handbook for Companies Doing Business in Emerging Markets. Washington, DC, 2007. https://www.ifc.org/wps/wcm/connect/affbc005-2569-4e58-9962-280c483baa12/IFC_StakeholderEngagement.pdf?MOD=AJPERES&CVID=jkD13-p

13.

Albritton

, Edmunds

, Thomas

, et al. Implementation Guide Number 1: Engaging Stakeholders to Improve the Quality of Children's Health Care. Rockville, MD: Agency for Healthcare Research and Quality, 2014.

14.

Carr

, Howells

, Chang

, et al. An integrated approach to stakeholder engagement. Healthc Q, 2009; 12(Spec No Ontario):62–70.

15.

Thom

, Jackson

, Leekha

, et al. Risk factors for surgical site infection following colorectal surgery. Open Forum Infect Dis, 2016; 3.

16.

Ray

, Miller

. Strengthening stakeholder-engaged research and research on stakeholder engagement. J Comp Eff Res, 2017; 6:375–389.

17.

Feldstein

, Glasgow

. A practical, robust implementation and sustainability model (PRISM) for integrating research findings into practice. Jt Comm J Qual Patient Saf, 2008; 34:228–243.

18.

Zerhouni

. Translational and clinical science—Time for a new vision. N Engl J Med, 2005; 353:1621–1623.

19.

Forsythe

, Ellis

, Edmundson

, et al. Patient and stakeholder engagement in the PCORI pilot projects: Description and lessons learned. J Gen Intern Med, 2016; 31:13–21.

20.

Hoffman

, Montgomery

, Aubry

, Tunis

. How best to engage patients, doctors, and other stakeholders in designing comparative effectiveness studies. Health Aff (Millwood), 2010; 29:1834–1841.

21.

Israel

, Schulz

, Parker

, Becker

. Review of community-based research: Assessing partnership approaches to improve public health. Annu Rev Public Health, 1998; 19:173–202.

22.

Wallerstein

, Duran

. Community-based participatory research contributions to intervention research: The intersection of science and practice to improve health equity. Am J Public Health, 2010; 100(Suppl 1):S40–46.

23.

Aimaq

, Akopian

, Kaufman

. Surgical site infection rates in laparoscopic versus open colorectal surgery. Am Surg, 2011; 77:1290–1294.