Risk factors: As per the NNIS system, it is well known that SSI is the second most common postoperative infection following caesarean deliveries, with an incidence ranging from 3 to 15%. In India, as per a study conducted at Lady Hardinge Medical College New Delhi, the infection rate was 24.2%. In our study, the approximate (excluding those lost to follow up) incidence of SSI in our hospital during one and a half years was 1.1%, which is much lower than the general incidence rate in various other studies done in India. In a study done by Anjum et al in 2017, at a tertiary care centre Tumkur in Karnataka, the SSI rate following a lower segment caesarean section (LSCS) was 16%, and another done by Priayanka Dahiya et al at a first referral unit in India in 2016, it was 9%. Studies done in the USA reported an incidence of 3.15% [18].
Based on studies done earlier, certain risk factors were identified as causative for SSI, namely obesity, diabetes mellitus (DM), HTN, PROM, emergency caesarean deliveries, chorioamnionitis, longer operative time, poor antimicrobial prophylaxis. Regarding the effect of maternal age, there have been controversial results in various studies although with increasing age due to the senescence of cells and poor host immune response mechanisms the risk of SSI is proposed to increase. In this study, the majority of cases (55%) of SSI were above 25years age and this was statistically significant. This was similar to studies conducted by Anjum et al and Wloch C.et al (2012), while a study was done by Olicer Ezechi et al (2009) in Nigeria showed 75% SSI cases below 25years age, replicating the outcomes from the study by Ashish Pathak et al (2017) in rural India [8-10]. A few have also reported an absence of any association between maternal age and SSI [3, 18].
Host related and obstetric risk factors found to have a statistically significant association for SSI in this study were obesity and multiparity (p = 0.013). Obesity as an independent risk factor correlated with studies done by Anjum et al (2017), Fathia et al (2012), and C Wloch et al (2012) (BMI>35kg/m2, p <0.01) [5, 7, 8]. Fathia et al reported an increase in SSI with BMI>35kg/m2, p= 0.005 [7]. Multiparity was also found to be a significant risk factor in the studies by Hansa Dhar et al (2014) and Filbert J Mpogoro et al (2012) [11, 23]. Majority of patients in our study were multiparous, moreover among them, those with previous h/o LSCS had an added risk, and this may have confounded the results. Among those with comorbidities in our study, hypertensive disorder of pregnancy and DM were highly significant in causation with a p-value <0.001 which correlated with studies by Anjum et al (2017), Rita Andersen (2011) [5, 12]. In the study by Rita Andersen among diabetic women obesity increased the risk by more than two-fold. In hypertensive disorders of pregnancy, the risk can be attributed to the increased rate of surgical intervention in preeclampsia, eclampsia which is supplemented by generalised oedema, and hypoalbuminemia [12]. The results of the study done by Bhadauria AR et al stated was found to be more among the patients having a pre-existing illness such as anaemia 21.13%, obesity 13.62%, DM 8.45% and HTN 4.69% (Table 4) [13].
Emergency surgeries, in general, are found to have a higher risk due to lack of adequate time for action of antibiotics, second stage labour, h/o PROM among them, and the possibility of improper preoperative preparation, that was also found significant in this study with p<0.001 among the emergency LSCS. Several other studies observed higher SSI rates in the emergency LSCS than elective cases [5, 14-18].
In this study, the reduced incidence of SSI with longer duration of PROM (more than 12hours), can be attributed to the early initiation and prolonged course of injectable antibiotics (continued for 5 days postoperatively, unlike for other cases with PROM <12 hours where postoperatively injectable preparations are given only for 2 days). This was following the study by Fathia et al, that had more SSI rate among those with PROM < 8hours, but it was not statistically significant, whereas Hansa Dhar et al and Olicer C Ezechi et al, showed a significant association with PROM (p<0.001) [7, 9, 11].
Surgical aspects:
Longer duration of surgery, mostly beyond 1 hour carried a significant association in our study (p=0.001), and also in the studies done by Fathia et al, Wloch C and Filbert J Mpogoro [7, 8, 23]. Martin et al study revealed, an operative time longer than 75th percentile increased the risk of SSI by 1.84 times [22]. The probable etiologies for an increase in SSI risk being: complicated surgery, inadequate tissue concentration of antibiotic, tissue trauma, breach of sterile technique, increased blood loss, and prolonged exposure to environmental pathogens (Table 5).
Skin closure by subcuticular sutures was found to have a higher risk of wound infection compared to mattress sutures, but this may have been confounded by the fact that mattress sutures were used for those with an obvious risk factor like obesity, uncontrolled DM or eclampsia, while the majority were sutured as subcuticular sutures. Staples were not used for any case. A Cochrane review has thus concluded that there is insufficient data on standardisation of any particular technique or material for LSCS wound closure.
Microbiological profile:
Out of the 52 swabs collected, 73% (38 cases) showed growth of an organism, while 27% showed absence of any microbial growth. Of the 38 cases, 52% showed growth of Staph aureus that was the same as the study by Fathia et al where Staph aureus comprised 50.4% of the growth while studies by Wloch C et al and Anjum et al showed 40% and 43% growth of Staph aureus among SSI [5, 7, and 8]. Among the 19 cases with Staph aureus growth, 60% constituted MRSA. Another study by Bhattacharya et al also showed the highest culture of Staph aureus (35%) among the SSI, (of which MRSA was 26%), others being E. coli (20%), Klebsiella (18%), Pseudomonas (8%) and Acinetobacter (7.5%) [19]. In our study, the other organisms isolated included Diphtheroids (12%), E. coli (10.5%), Klebsiella (7%), Pseudomonas (7%), and Acinetobacter (2%). The main gram-negative offenders in our study were E. coli, followed by Klebsiella and Pseudomonas, this was similar to that in studies done by Bhattacharya and Anjum et al [5, 19]. No growth of Streptococci was obtained.
The highest sensitivity was observed for aminoglycosides like Gentamicin and glycopeptides like Linezolid and Vancomycin. Methicillin-resistant Staphylococcus aureus showed a higher sensitivity for Clindamycin and Linezolid, while MSSA were more sensitive to Gentamicin and Clindamycin. Gentamicin was effective against E. coli and Pseudomonas. In a study conducted in Mumbai Lilani SP et al in 190 patients, Pseudomonas showed high resistance to Gentamicin [20]. In our hospital, gentamicin was rarely used as a postoperative antimicrobial, unless indicated so lesser were the chances of antibiotic resistance. Unlike other studies, Diphtheroids being a component of normal skin flora were the second most common contaminating microbe cultured, they showed relatively high sensitivity towards Vancomycin.
As stated in several other studies, the most effective antibiotics for gram-positive infections were Linezolid and Vancomycin (90.9% and 81.8% bacteria). However, because of emerging drug resistance, the use of these antibiotics for empirical therapy cannot be recommended. These are preferable to be used only in life-threatening infections or resistance to other commonly used drugs. Among the gram-negative organisms, Ceftazidime or Gentamicin could be used as second-line therapy.
As per the study by Das R et al, for gram-negative infections, a combination of Piperacillin + Tazobactam or Cefoperazone + Sulbactam and Amikacin [21]. It was found that 60% of staphylococcal infections were methicillin-resistant. These MRSA also showed an increasing trend towards resistance to higher antibiotics like Carbapenems. There was no Vancomycin-resistant Staphylococcus aureus (VRSA) isolated. In a study from Northern India, two (0.25%) VRSA strains were reported [20]. Gram-negative organisms were also noted to have developed resistance to broad-spectrum antibiotics like Piperacillin (Table 6).