American Journal of Infection Control 42 (2014) e11-e15

Contents lists available at ScienceDirect

American Journal of Infection Control

American Journal of Infection Control

journal homepage: www.ajicjournal.org

Major article

Incidence and factors associated with surgical site infections in a teaching hospital in Ujjain, India Ashish Pathak DCH, DNB (Pediatrics), PhD a, b, c, Erika A. Saliba RN, MSc a, *, Shailendra Sharma MS d, Vijay Kumar Mahadik DCH e, Harshada Shah MD f, Cecilia Stålsby Lundborg PhD a a

Global Health/IHCAR, Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden Department of Paediatrics, Ruxmaniben Deepchand Gardi Medical College, Ujjain, Madhya Pradesh, India Department of Women and Children’s Health, International Maternal and Child Health Unit, Uppsala University, Uppsala, Sweden d Department of Surgery, Ruxmaniben Deepchand Gardi Medical College, Ujjain, Madhya Pradesh, India e Ruxmaniben Deepchand Gardi Medical College, Ujjain, Madhya Pradesh, India f Department of Microbiology, Ruxmaniben Deepchand Gardi Medical College, Ujjain, Madhya Pradesh, India b c

Key Words: Health care-associated infections Antibiotics Surveillance Risk factors Staphylococcus aureus Pseudomonas aeruginosa Rural

Background: Surgical site infections (SSI) are among the most commonly reported health care-associated infections; however, there is a paucity of data on SSI from India. This study aimed to determine the incidence of SSI and explore its associated factors at a teaching hospital in India. Methods: Direct and indirect surveillance methods, based on Centers for Disease Control and Prevention guidelines, were used to define SSI. Patients were followed up for 30 days postsurgery. Prescribing and resistance data were collected. Results: The SSI rate among the 720 patients investigated was 5%. Risk factors for SSI identified were as follows: severity of disease (P ¼ .001), presence of drains (P ¼ .020), history of previous hospitalization (P ¼ .003), preoperative stay (P ¼ .005), wound classification (P < .001), and surgical duration (P < .001). Independent risk factors identified included wound classification (odds ratio ¼ 4.525; P < .001) and surgical duration (odds ratio ¼ 2.554; P ¼ .015). Most patients (99%) were prescribed antibiotics. Metronidazole (24.5%), ciprofloxacin (11%), and amikacin (9%) were the most commonly prescribed antibiotics. Most commonly isolated bacteria were Staphylococcus aureus (n ¼ 14), of which 34% were methicillin-resistant Staphylococcus aureus, and Pseudomonas aeruginosa (n ¼ 6), which showed resistance to ceftazidime (70%), ciprofloxacin (63%), and gentamicin (57%). Conclusion: Incidence of SSI at the hospital was lower than reported in many low- and middle-income countries, although higher than reported in most high-income countries. Targeted implementation strategies to decrease incidence of preventable SSI are needed to further improve quality and safety of health care in this hospital and similar hospitals elsewhere. Copyright Ó 2014 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.

* Address correspondence to Erika A. Saliba, RN, MSc, Global Health/IHCAR, Department of Public Health Sciences, Karolinska Institutet, Tomtebodavägen 18A, SE 171 77 Stockholm, Sweden. E-mail address: [email protected] (E.A. Saliba). A.P. and E.A.S. contributed equally to this paper. This work made part of a master’s degree thesis in Global Health for Erika A. Saliba at Karolinska Institutet. Supported by Vetenskapsrådet (Swedish Research Council) and Asia Link. E.A.S. received a Strategic Educational Pathways Scholarships (STEPS) scholarship for master's degree studies at Karolinska Institutet. STEPS is supported, in part, by the European Social Fund’s Operational Programme II Cohesion Policy 2007-2013, “Empowering People for More Jobs for and a Better Quality of Life.” Conflicts of interest: None to report.

Despite major advances in infection control interventions, health care-associated infections (HAI) remain a major public health problem and patient safety threat worldwide.1,2 The global estimated prevalence of HAI, at any given time, approximates 1.4 million.3 Surgical site infections (SSI) are among the most commonly reported HAI.4 Incidence varies widely across countries and surgical procedures; however, it is estimated to occur in at least 2% of surgeries.5 In low- and middle-income countries (LMIC), SSI incidence may be approximately up to 4 times higher than in highincome countries.1 SSI involve a complex relationship among several factors: microbial, patient, surgical, and environmental.6

0196-6553/$36.00 - Copyright Ó 2014 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajic.2013.06.013

e12

A. Pathak et al. / American Journal of Infection Control 42 (2014) e11-e15

They are also dependent on the health care facility, procedures performed, and care level.1 For individual patients, SSI are associated with longer hospitalization; pain; discomfort; delayed wound healing; prolonged or permanent disability; and, in worst cases, death.2 Additionally, SSI place significant economic burden on health system and patient finances and resources because of lengthier hospitalizations and increased cost of treatment.5 In India, the risk of acquiring SSI is high (range, 4%-30%).7 Nonetheless, prevention and control of HAI is not prioritized and antibiotic (AB) resistance is an ever-growing problem.8 Moreover, surveillance data are still scanty.1,9 By forming evidence-based strategies to diminish rates of preventable SSI and their adverse effects, one can increase patient safety, circumvent additional health care costs, and ameliorate health care quality.5 This study sought to determine the incidence of SSI and explore its associated factors at Chandrikaben Rashmikant Gardi Hospital (CRGH), Ujjain, India. METHODS Study setting The study was conducted in 90-bed general surgery wards of CRGH, a 570-bed teaching hospital associated with Ruxmaniben Deepchand Gardi Medical College (RDGMC). The obstetrics and gynecology; ear, nose, and throat; and orthopedic wards were not included.

disc screen test and 6 g/mL oxacillin in Mueller-Hinton agar supplemented with NaCl (4% wt/vol; 0.68 mol/L) according to CLSI guidelines.13 Statistical analyses Data were entered in EpiData Software (version 3.1, EpiData Software Association, Odense, Denmark); SPSS Statistics (version 20.0; SPSS Inc, Chicago, IL), and Stata (version 10.0; Stata Corp, College Station, TX) were used for analyses. Frequencies and percentages were determined for binary and categorical variables. Range, mean, standard error (SE), and standard deviation were calculated for continuous variables. Cumulative incidence rate of SSI was also calculated. The relationship between risk factors and SSI was explored using Pearson c2 test. In bivariate analyses, a P value of < .1 was considered significant for entry into a backward multivariate logistic regression model, with SSI as outcome variable. Adjusted odds ratios and their respective 95% confidence intervals (CI) were calculated. A P value of < .05 was considered significant in the final model. No associations were made from variables with cell counts less than 5, which were excluded from the final logistic regression model. RESULTS Demographics

Study participants All patients admitted between October 2010 and August 2011 were included prospectively in the survey. All nonsurgical cases and patients not undergoing surgery at CRGH were not included. Ethical permission was obtained from the Ethics Committee of RDGMC (approval number 114/2010).

Of 1,765 admitted patients, 41% underwent surgery. Three patients did not undergo surgery at CRGH. The final cohort therefore included 720 patients: 76% male and 24% female. Ages ranged from 4 months to 90 years (mean  standard deviation: 40.10  21.23). Fifteen patients died during the study duration (15/720; 2%), out of which 3 patients had contracted SSI.

Study design

Incidence of SSI

The US Centers for Disease Control and Prevention (CDC) surveillance methods for SSI was used.10 A trained study assistant conducted indirect surveillance by acquiring patient information using a form containing SSI risk factors. The study assistant also inspected all surgical sites at time of dressing change 24 to 48 hours postsurgery (direct surveillance), took swabs of suspected SSI and sent them for analysis, and completed postdischarge surveillance. For postdischarge surveillance, patients were asked to return for follow-up 30 days postdischarge at the hospital’s surgical outpatient clinic. If this did not occur, patients were contacted by mobile phone, and, if an SSI was suspected, they were asked to return to CRGH to confirm the diagnosis. CDC’s National Health Safety Network criteria were used for diagnosing SSI.10 Details of patients’ AB prescriptions were collected and presented in the form of drug utilization 90% (DU90%).11

SSI occurred in 5% (95% CI: 3-7) of patients; all 5% (n ¼ 36) were confirmed by visual impression, and 3% (n ¼ 22) further confirmed by CDC definition through a positive culture from infection site. CDC-confirmed SSI were further classified into superficial incisional primary (1.5%; n ¼ 11), superficial incisional secondary (0.8%; n ¼ 6), deep incisional primary (0.4%; n ¼ 3), or deep incisional secondary (0.1%; n ¼ 1).

Swab sampling and laboratory methods Swab samples were plated on blood agar and MacConkey agar medium. Standard conventional microbiological methods were used to identify pathogenic bacteria.12 The Kirby-Bauer disc-diffusion method on Mueller-Hinton agar plates was used for AB susceptibility testing. Disc strengths were as recommended by the Clinical and Laboratory Standards Institute (CLSI). CLSI interpretative criteria for susceptibility and resistance testing were used.13 Intermediate susceptible isolates of gram-negative bacteria were considered resistant in calculations. For Staphylococcus aureus isolates, screening for methicillin resistance was done using cefoxitin

Risk factors associated with SSI Few patients suffered from chronic diseases, including cardiac, renal, or hepatic disease; diabetes; or tuberculosis. Almost half were smokers (n ¼ 298), of whom 98% were male and 2% female. Few patients (1.7%) were immunosuppressed at time of surgery. Severity of disease, measured using American Society of Anesthesiologists (ASA) score, ranged from healthy (class I) to severe systemic disease, which is a constant threat to life (class IV). The proportion of patients in each class from class I to class IV was 71%, 22%, 6%, and 0.4%, respectively. Full description of patient risk factors associated with SSI is displayed in Table 1. The mean duration of preoperative stay was 4 days (SE  0.21), whereas the mean duration of postoperative stay was 8 days (SE  0.29). For SSI patients, the mean preoperative stay was 6.6 days (SE  1.44) and 19 days postoperatively (SE  2.5). Some patients (6%) had a history of previous hospitalization (hospitalization/s maximum 2 weeks prior admission). Infection prior to surgery was detected in 17% of patients. The majority showered and had hair removed preoperatively, mostly by shaving. Most patients

A. Pathak et al. / American Journal of Infection Control 42 (2014) e11-e15

e13

Table 1 Descriptive data and bivariate analyses of patient risk factors associated with SSI in Ujjain, India Total number of participants: 720 %total (N ¼ 720) Sex Male Female Age (y) 18 19-35 36-50 51-65 66 Chronic diseasesy Yes No Tuberculosis Yes No Renal disease Yes No

%SSIyes (n ¼ 36)

%SSIno (n ¼ 684)

P value .533

76.2 23.8

5.3 4.1

94.7 95.9

18.5 26.5 22.6 21.8 10.6

0.0 4.7 7.4 6.4 6.6

100.0 95.3 92.6 93.6 93.4

11.5 88.5

6.0 4.9

94.0 95.1

3.3 96.7

0.0 5.2

100.0 94.8

0.7 99.3

0.0 5.0

100.0 95.0

.768*

.649

NA

NA

%total (N ¼ 720)

%SSIyes (n ¼ 36)

1.8 98.2

7.7 5.0

92.3 95.0

3.1 96.9

4.6 5.0

95.4 95.0

3.6 96.4

11.5 4.8

88.5 95.2

41.4 58.6

6.7 3.8

93.3 96.2

1.7 98.3

0.0 5.1

100.0 94.9

93.5 6.5

4.3 14.9

95.7 85.1

Diabetes Yes No Cardiac disease Yes No Hepatic disease Yes No Smokerz Yes No Immunosuppression Yes No ASA score Class I & class II Class III & class IV

%SSIno (n ¼ 684)

P value .653

.921

.119

.077

NA

.001

ASA, American Society of Anesthesiologists; SSI, surgical site infection. NOTE. NA ¼ unable to carry out statistical analysis because cell count ¼ 0. *Participants  18 years of age excluded from analysis (no cases of SSI). y Includes a combination of chronic diseases: cardiac, renal, and hepatic disease and diabetes and tuberculosis. z Includes present smokers (does not include exsmokers).

were supplemented with oxygen during surgery, with few becoming hypoxic intraoperatively. Drains were inserted in 19% of cases. Full description of surgical risk factors associated with SSI is displayed in Table 2. The majority of surgeries (96%) was elective. Most patients underwent major surgery (79%), and many were clean (23%) or cleancontaminated (65%). Only a few were contaminated (7.5%) or dirty (5%). Most surgeries lasted less than 1 hour (79%). Surgeries performed varied from simple dermatologic procedures such as excisions and incision and drainages to major abdominal, renal, pulmonary thoracic, or genitourinary surgeries, as well as surgeries for gangrene removal. The most common procedures included the following: hernioplasties, herniotomies, exploratory laparotomies in the lower gastrointestinal tract, and appendectomies (45%), followed by dermatological procedures (27%). Least common were those to remove gangrenous tissue (1.5%). SSI were most common in gangrenous (27%), renal (12.5%), genitourinary (5.8%), and lower gastrointestinal tract (4.4%) surgeries. Before adjusting for age and sex, SSI was significantly associated with ASA classes III and IV (P ¼ .001), longer preoperative stay (P ¼ .005), previous hospitalization (P ¼ .003), contaminated and dirty surgeries (P < .001), surgeries lasting 1 hour or more (P < .001), and drain insertion (P ¼ .020). After adjusting for age and sex, results from the multinomial logistical regression model showed that wound classification and surgical duration are independently associated risk factors for SSI development (odds ratio ¼ 4.525 [95% CI: 2.014-10.168], P < .001, and odds ratio ¼ 2.554 [95% CI: 1.199-5.443], P ¼ .015, respectively). AB prescription ABs were prescribed to almost all admitted patients (99%). Onethird were prescribed 1 AB (commonly metronidazole), 25% were prescribed a combination of 2 ABs (commonly ciprofloxacin with metronidazole), and 17% received a combination of 3 ABs (commonly ciprofloxacin with metronidazole and amikacin). The DU90% included the following: metronidazole (24.5%), ciprofloxacin (11%), amikacin (9%), cotrimoxazole (9%), doxycycline (7.5%), cefotaxime (7%), gentamycin (7%), norfloxacin (7%), lincomycin

(6%), and coamoxyclav (5%). The mean duration of AB therapy was 4.5 days (95% CI: 4.44-4.60), and the median was 4 days. AB susceptibility pattern of SSI isolates From the 44 swabs taken, 22 bacteria were isolated, the most common being S aureus (n ¼ 14), followed by Pseudomonas aeruginosa (n ¼ 6), and Escherichia coli (n ¼ 2). Methicillin-resistant S aureus constituted 34% of all S aureus isolates and showed resistance toward coamoxyclav (82%), ciprofloxacin (40%), and amikacin (30%). None of the S aureus isolates showed resistance to vancomycin. P aeruginosa isolates showed resistance toward cetazidime (70%), ciprofloxacin (63%), gentamicin (57%), amikacin (29%), and imipenam (8%). DISCUSSION To our knowledge, this is the first SSI surveillance study in a central rural area in India, which describes incidence and associated risk factors of SSI using CDC definitions and 30-day follow-up surveillance. Previous Indian studies have shown SSI rates ranging between 4% and 30%.7 These high rates could be due to lack of infrastructure, human resources, clean water, poor sanitation, overcrowding, and often a general disinterest in basic infection control by health care staff.14 This study reports a rather low SSI incidence (5%), lower than reported in Gujarat (8.95%)15 and Mumbai (12.72%),16 India, as well as Iran (17.4%),17 Egypt (17%),18 and Pakistan (7.3%).19 However, in Dehradun, India, a 5% incidence was also reported.20 Although 5% is considered rather low for LMIC, it remains higher than in many high-income countries where SSI rates typically range between 1.2% and 5.2%.1 Our low SSI incidence could be due to multiple factors; however, it is strongly believed that the young patient age (most patients fell within the 19-35 year age group), the low prevalence of chronic disease in this cohort, and the elective nature of the majority of surgeries (96%) could have resulted in low SSI rates. This is also reflected in the patient’s ASA scores, with 93.5% falling within classes I and II. Another reasonable explanation is that, because patients were contacted by mobile phone during the 30-day follow-up, they may

e14

A. Pathak et al. / American Journal of Infection Control 42 (2014) e11-e15

Table 2 Descriptive data and bivariate analyses of surgical risk factors associated with SSI in Ujjain, India Total number of participants: 720 %total (N ¼ 720) Days hospitalized preoperatively* 0-3 4-7 8-14 15 Previous hospitalizationz Yes No Preoperative infectionx Yes No Duration of surgery (min)

Incidence and factors associated with surgical site infections in a teaching hospital in Ujjain, India.

Surgical site infections (SSI) are among the most commonly reported health care-associated infections; however, there is a paucity of data on SSI from...
186KB Sizes 0 Downloads 0 Views