Post Caesarean Wound sepsis and associated factors among patients attending a rural regional referral hospital in Western Uganda: A cross-sectional study

Background : Post-caesarean wound sepsis is among the most common problem for patients who undergo caesarean section. It remains a common and widespread problem contributing to morbidity and mortality; this could be due to an increase in antimicrobial resistance. Determining the burden of wound sepsis and common bacterial pathogens can provide solution to prevent incidence and establish microbiological mapping. Aim: To determine prevalence, identify factors, common bacterial pathogens from post-caesarean wounds and antibacterial susceptibility pattern at Hoima Regional Referral Hospital. Methods : A cross-sectional study was conducted among post-caesarean mothers attending Hoima Regional Referral Hospital. Consecutive enrolment of 303 participants who consented to participate was done. Structured questionnaires were used to collect data on associated factors and wound swabs were done prior to bacterial culture. Antibacterial susceptibility pattern of isolated bacterial pathogens was determined by Kirby Bauer disc diffusion method. Data was analyzed using Stata 14.2. Results : The wound sepsis rate was 16.8%. Being educated, multiple vaginal examination, hygiene, previous caesarean sections and HIV seropositivity were all significantly and positively associated with post-caesarean wound sepsis (p<0.05). The most implicated bacteria was Staphylococcus aureus and was most susceptible to ciprofloxacin. Resistance was most exhibited against ciprofloxacin, gentamycin, ceftriaxone and cotrimoxazole especially by coliforms. Hospital educated, hygiene, caesarean


Abstract
Background : Post-caesarean wound sepsis is among the most common problem for patients who undergo caesarean section. It remains a common and widespread problem contributing to morbidity and mortality; this could be due to an increase in antimicrobial resistance. Determining the burden of wound sepsis and common bacterial pathogens can provide solution to prevent incidence and establish microbiological mapping.
Aim: To determine prevalence, identify factors, common bacterial pathogens from postcaesarean wounds and antibacterial susceptibility pattern at Hoima Regional Referral Hospital.
Methods : A cross-sectional study was conducted among post-caesarean mothers attending Hoima Regional Referral Hospital. Consecutive enrolment of 303 participants who consented to participate was done. Structured questionnaires were used to collect data on associated factors and wound swabs were done prior to bacterial culture. Antibacterial susceptibility pattern of isolated bacterial pathogens was determined by Kirby Bauer disc diffusion method. Data was analyzed using Stata 14.2.
Results : The wound sepsis rate was 16.8%. Being educated, multiple vaginal examination, hygiene, previous caesarean sections and HIV seropositivity were all significantly and positively associated with post-caesarean wound sepsis (p<0.05). The most implicated bacteria was Staphylococcus aureus and was most susceptible to ciprofloxacin. Resistance was most exhibited against ciprofloxacin, gentamycin, ceftriaxone and cotrimoxazole especially by coliforms.
Conclusions. The rate of caesarean wound sepsis is high at Hoima Regional Referral Hospital . Being educated, multiple vaginal examination, hygiene, previous caesarean sections and HIV seropositivity are possible risk factors for the condition. Staphylococcus aureus is the commonest organism isolated from exudates of septic wounds after caesarean section. Awareness amongst health workers and patients about these major factors is necessary so that management can be directed. Rational use of antibiotics by health workers is paramount to combat resistance in this setting.

Background
Post-cesarean wound sepsis is common among women following cesarean section and this can easily progress into septic shock if poorly managed [1,2]. Early diagnosis of postcesarean wound sepsis using sepsis biomarkers such as pro-calcitonin (PCT) is an important medical practice [3]. These biomarkers are hardly available in several health care units in developing countries and this has led to the development of a global campaign on development of appropriate guidelines for successful management of sepsis [4,5]. Successful management is crucial to ensure that unnecessary prolonged activation of the immune response is inhibited [6][7][8]. Though cesarean sections are done under aseptic conditions, the risk of post-cesarean wound sepsis always exists which puts postcesarean wound sepsis among the most common nosocomial infections [9]. Globally, the rate of cesarean section is increasing rapidly [10] and the most common complication after cesarean section is wound sepsis, with an incidence of 3-15% [11]. Post-cesarean wound sepsis is also associated with maternal mortality rate of up to 3% especially in health units that have no facilities to carryout safe cesarean sections or treat postcesarean complications [12].
In sub-Saharan Africa, post-cesarean wound sepsis has been shown to be in the range of 1.7% to 10.4% showing that the condition is significant in the continent [13,14]. This has been attributed to poor accessibility to health care services which is below 3% and thus resulting in poor postnatal follow up and wound care [15]. In addition, post-cesarean wound sepsis in sub Saharan Africa has been associated with poverty, environmental pollution, poor preoperative care, malnutrition, anemia, wound contamination, poor antibiotic selection and poor immunity [16]. In East Africa, the rate of cesarean sections is below 40%, which is a higher rate [17] compared to the recommended rate of cesarean section as per the WHO which considers that the best cesarean section rate is between 10-15% [18]. This increases the risk of post-cesarean wound sepsis.
In Uganda, a study conducted in Jinja showed that the major pathogenic bacteria associated with post-operative sepsis are; Staphylococcus aureus, coliforms , Proteus mirabilis, Pseudomonas aeruginosa, Klebsiella pneumoniae and Enterobacter spp [19]. In addition, the study showed that these pathogens were highly resistant to ampicillin, amoxicillin and chloramphenicol. A study conducted at Mulago National Referral Hospital in Uganda showed that majority of these septic wounds have pathogens that are methicillin resistant thus posing a major healthcare challenge due to the limited class of drugs available in several healthcare centers in Uganda [20]. At Hoima Regional Referral Hospital (HRRH) there was a high turnover of patients with eight to ten cesarean sections done in a day (maternity theatre register). It was also noted that theatres were being shared by other surgical teams and that the patients with post-cesarean wound sepsis were not isolated from the rest of the patients in ward, according to the findings during an onsite visit to the Hospital.
Several associated factors including patient related factors, hospital factors and obstetric factors have made management of post-cesarean wound sepsis challenging, leading to an increase in hospital stays [21]. Hence, thorough identification of the bacterial pathogens and the associated factors of post-cesarean wound sepsis is important for developing proper protocols to reduce its burden and complications.

Study design
We performed a descriptive and analytical cross-sectional study and we involved laboratory investigations to isolate the bacterial pathogens and antibacterial susceptibility pattern. The outcome variables were post-cesarean sepsis. We also assed predictors such as socio-demographic and clinical factors for the study participants.

Study site and setting
The study was conducted in the postnatal ward at Hoima Regional Referral Hospital (HRRH) between the months of July to September 2018. HRRH is a public hospital located in Hoima District with in coordinates; 01 24N, 31 18E and is approximately 230 km by road from Kampala (Capital city of Uganda). The major tribe is Banyoro and the main occupation is animal husbandry and crop farming.
Hoima Regional Referral Hospital is a well-established hospital and it offers both in-patient and out-patient services with a range of departments and clinics, including General Surgery, Orthopedics, Obstetrics and Gynecology and Internal Medicine. The hospital is well equipped with a bed capacity of 400. The Obstetrics and Gynecology Department of Hoima Regional Referral Hospital has four specialists, one Resident doctor, five intern doctors and 13 midwives. The obstetrics and gynecology department has 110 beds. The post-natal ward for post-cesarean mothers is located in the new block. The new block has about 60 beds which are always fully occupied with some patients lying on the floor.
According to the hospital records (Maternity Theatre register), Hoima Regional Referral Hospital performs approximately 8-10 cesarean sections per day. These are done in two theatres that are shared by other surgical teams. The hospital had a range of 10-20 vaginal deliveries per day and also offers antenatal and postnatal services. The main laboratory of Hoima Regional Referral Hospital consists of the following sections: hematology and blood bank, chemistry, parasitology and microbiology. It was composed of 20 staff members and these include three specialists, two laboratory technologists, eight laboratory technicians, six laboratory assistants, and one laboratory attendant. The microbiology laboratory was operated by one laboratory technologist, one laboratory technician, one laboratory assistant and one laboratory attendant. It was well equipped to carry out culture and sensitivity and other microbiological tests. Some of the equipment found in this laboratory were; autoclave, incubator, microscope, hot air oven, refrigerator, safety cabinet and gas cylinder.

Study population
The study population were mothers who have delivered by cesarean section at Hoima Regional Referral Hospital during the period of the study.

Inclusion Criteria
All adult and emancipated mothers (on ward or re-admitted) who delivered by cesarean section at Hoima Regional Referral Hospital during the study period.

Exclusion Criteria
Cesarean section mothers from other health units that were referred to Hoima regional referral hospital were excluded due to limited access to their medical records. Mothers who would have had a re-exploration due to cesarean section complications other than suspected sepsis, as well as those who were in their early puerperium and those who reported after six weeks, were also excluded.

Sample size
The sample size for this study was 303 and was computed using modified Kish Leslie formula (Daniel, 1999).

Sampling technique
Consecutive enrollment of participants who consent to participate in the study. This was carried out on a daily basis until required sample size.

Data collection instruments
Structured investigator-administered pre-tested questionnaire was used for each participant to collect information on socio-demographic and known factors that may be related to the development of post-cesarean wound sepsis in each patient including obstetric factors, hospital factors and health factors.
A detailed history was elicited (English), translated where necessary for women who did not understand English; and physical examination was performed. Presence or absence of post-cesarean wound discharge (exudate) was noted. Swabs for mothers with discharge were taken and cultured in the laboratory according to standard clinical laboratory guidelines. Susceptibility testing was carried out according to Kirby Baur diffusion methods.

Sample collection and transportation
Patients with post-cesarean wound sepsis who met the inclusion criteria of the study and were, educated, counseled about the study and then consented to participate. They were requested to allow history taking and physical examination and when a patient had postcesarean wound discharge (Exudate), a sample was taken for microbiological analysis.
Using sterile swab sticks, two samples from each participant were collected by gently rubbing the sterile swab sticks in the infected site (wound depth) using aseptic technique and immediately replaced inside the swab sticks case. The sterile swab stick was labeled with each participant's study number and transported to Microbiology laboratory immediately for processing, and in case of any delay, the sample was stored aerobically in the refrigerator at 4-8⁰C.

Isolation
The collected samples were inoculated on blood agar, chocolate agar, MacConkey agar and mannitol salt agar. They were then incubated both aerobically and anaerobically at 37ᴼC for 24-48hrs. Colony morphology was observed according to shape, size, elevation and margin and surface characteristics.

Direct Gram Microscopy
A direct smear was made for Gram stain; a drop of sterile normal saline was added at the center of a clean dried glass slide and the swab containing the sample rolled in the drop of normal saline spreading it on the glass slide in a circular motion to make a thin smear of the size of a fifty shilling coin. The smear was allowed to air-dry and then heat-fixed by passing it at least three times over a Bunsen flame. The slide was placed on the staining rack and flooded with crystal violet solution for 60 seconds, washed with clean water and covered with Lugol's Iodine (a mordant) and then allowed to act for a minute.
The slide was again washed in clean water and then decolorized with 50% acetonealcohol under slow running tap water until a faint pink color was observed or no more color tend to flow from the smear. The process of decolorizing did not exceed 30 seconds.
After decolorizing, the slide was washed in clean water and counterstained with neutral red solution. The slide was then washed in clean water; air-dried and observed under the microscope with x100 objective lens (oil immersion lens). Gram-positive bacteria was observed as blue or purple color and Gram-negative as red or pink color. Also, the morphology and shape of the bacteria was used to identify whether they are cocci, diplococcic, cocci in chains, clusters, and whether they are rods in appearance. Pus cells were also observed in the direct Gram-stained slide.

Cultural characteristics
The colony morphological characteristics of the bacterial isolates was observed as follows; color, margin, mucoid, texture, and hemolysis on blood agar medium, among others. This helped in determining the characteristics of the colonies of the bacteria on culture media such as Lactose or non-lactose fermenters on MacConkey agar and type of hemolysis (alpha, beta, and gamma hemolysis) on blood agar.

Biochemical tests
The isolates were identified using the conventional biochemical tests such as API (Analytical Profile Index) which depends on the Gram's stain reaction of the isolates (Gram-positive or Gram-negative bacteria). Other biochemical tests included catalase, optochin, bacitracin, coagulase, indole, citrate utilization, urea utilization, triple sugar iron agar fermentation, MR-VP test and oxidase as described below:

Catalase test
The Catalase Test was carried out to differentiate between Streptococcus and Staphylococcus species and this was done according to standard methods [22] . A drop of 3% hydrogen peroxide was added to a loop full of the test organisms. Presence of bubbles indicated catalase activity. Streptococcus species was catalase positive while Staphylococcus species was catalase negative.

Indole test
The Indole Test was carried out according to the method described by Cheesbrough [22] to determine the ability of the isolate to degrade amino acid tryptophan and produce tryptophanase enzyme. A 1% tryptophan broth in a test tube was inoculated with 7 days isolate and incubated at 37°C for 48 hours. After 48 hours, 1 ml of chloroform was added to the broth. The test tube was shaken gently, and 2.1 ml of Kovac's reagent was added and again shaken gently. This was allowed to stand for 20 minutes. The formation of red coloration at the top layer, indicated a positive test, while a yellow coloration indicated negative result. Escherichia coli and Proteus are indole-positive.

Urease test
The Urease test was carried out according to the method described by Cheesebrough [22] to determine the ability of the bacteria to hydrolyse urea and produce ammonia and carbon dioxide. The test organism were inoculated into urease broth and incubated at 30°C for 72 hours. Purplish pink coloration of the medium indicated a positive reaction for Proteus and negative for other enterobacteria like Klebsiella and E. coli.

Citrate utilization
This was carried out by inoculating the test organism in test tube containing Simon's citrate medium and incubated for 24 to 72 hours. The development of deep-blue color after incubation was indicate a positive result. Klebsiella species are citrate-positive.

Triple sugar-iron test
Triple sugar iron test was carried out according to the method described by Cheesbrough [22]; the test determined the ability of the organism to ferment the three sugar component of the medium: glucose, lactose and sucrose. The medium contains a pH indicator (phenol red) and a detection system (thiosulphate and ferrous sulphate) for hydrogen sulphide (H 2 S). The medium was prepared as an agar slant. The test organism was inoculated by stabbing the medium using sterilized straight wire loop and the surface of the slope was also streaked with the test organism. The test was incubated at 37°C for 3 days. After incubation, gas production was determined by observing the cracking of the medium, and production of H 2 S was observed by the blackening of the butt (bottom) of the medium. The triple-sugar iron-agar aided in identification of Escherichia coli which ferments all three sugars and produce acid, turning the media into yellow color. Proteus species produces H 2 S which is indicated by black coloration of the media and fermentation at the butt of the tube.

Methyl red -Voges -Proskauer test (MR-VP)
Methyl red -Voges -Proskauer test (MR-VP) was carried out according to the method described by Cheesbrough [22]. It was used to determine the ability of the organisms to ferment glucose with production of acid. Five milliliters (5 ml) of MR-VP broth were inoculated with the test organism and incubated for 48 to 72 hours at 37°C. After incubation, 2 to 3 drops of methyl red test were added to 1ml of the broth. A red color signified a positive methyl red test, while yellow color signified a negative test. To what remained, five drops of 4% potassium hydroxide (KOH) were added followed by fifteen drops of 5% α -naphthol in ethanol. The development of red color within 1 hour indicates VP positive test while no color change indicated VP negative test. Escherichia coli is methyl red positive and voges-proskauer negative.

Coagulase test
It was used to identify Staphylococcus aureus which produces the enzyme coagulase. The rapid slide test was done by placing a drop of distilled water on each end of slide. Then a colony of the test organism (previously checked by Gram-staining) was emulsified in each of the drops to make two thick suspensions. A loopful of plasma was added to one of the suspensions (no plasma was added to the second suspension), and mixed gently.
Formation of clumps of the organisms within 10 seconds was indicative of a positive test while absence of these clumps was indicative of negative results.
For suspected Staphylococcus aureus isolates which turn negative for the rapid slide test, the test was done by emulsifying several isolated colonies of test organism in 1 ml of diluted rabbit plasma (1:5) dilution to give a milky suspension. The tubes were then incubated at 35°C in water bath for 4 hours. These were then examined at intervals of 1, 2 and 4 hours for clot formation by tilting the tube through 90°. If the test was still negative, the tube was left at room temperature overnight and examined again for Staphylococcus aureus that produced a delayed clot.

Oxidase test
The test was used in identification of organisms which produce the enzyme cytochrome oxidase. A filter paper soaked with the substrate tetramethyl-pphenylenediamine dihydrochloride was moistened with sterile distilled water. Using a glass rod, a colony of the test organism was smeared on the filter paper. The development of a blue-purple color within a 10 seconds was indicative of positive test while absence or formation of a blue-purple color after 10 seconds was considered negative. Pseudomonas species and Neisseria species are oxidase positive.

Susceptibility Pattern Determination (Kirby-Bauer disc diffusion technique)
The susceptibility pattern of the identified pathogens was determined by agar disc diffusion technique by Kirby-Bauer using Muller Hinton agar. The Muller Hinton agar was prepared according to the manufacturer's instructions under sterile conditions to avoid any contamination that may result. About 4-5 colonies of the organism were diluted in sterile peptone water and mixed and incubated for 4-5 minutes and its turbidity compared to a McFarland standard 0.5%.
A prepared Muller Hinton agar was dried in an incubator for at least 15 minutes and using a sterile glass rod or swab stick, the surface were smeared with the diluted organism of the peptone water, and using sterile forceps, ten different antibiotic containing discs were placed on the dried surface of the Muller Hinton agar containing the organism and then placed in the incubator at 37⁰C for 24 hours.
The diameter of a clear zone surrounding the antibiotic disc was measured in millimeters and compared to a standard antibiotic chart for measuring zones of inhibition. Zones of inhibition measured were recorded as susceptible (S), Intermediate (I), and Resistant I according to the standard chart.

Statistical analysis
Data on questionnaires was entered in Microsoft Excel version 2010, and then data from Excel was exported to IBM SPSS statistics version 23 as well as STATA 14.1 (Statacorp, USA Texas). Socio-demographic and clinical factors were summarized as means and medians, standard deviations and interquartile range (for continuous variables).
Proportions, percentages and frequencies were used for categorical variables using STATA 14.1.
Prevalence of post caesarian sepsis was summarized as frequencies and percentages. The factors associated with post-cesarean wound sepsis were assessed using binary logistic regression. For factors with odds ratio ≤ 0.2 at bivariate and those with biological plausibility were transferred for multivariate analysis. Bar charts were used to present susceptibility profile.

Quality control
The questionnaires was checked for completeness before collection to ensure valid data is obtained. Two independent microbiologists were involved in identification bacterial isolates. Additionally, standard operating procedures were adhered to.

Ethical considerations
Permission to conduct this study was obtained from the department of obstetrics and gynecology and post graduate directorate. Afterwards approval was obtained from Kampala International University Research Ethics Committee (KIU-REC).

Results
Socio-demographic characteristics of study participants attending Hoima Regional Referral Hospital Table 1. The median age was 23 years for both referred and non-referred study participants, with 92.2%married among the non-referred participants. Majority of the nonreferred study participants had no formal education (56%) compared to 24% among referred participants (p=0.001). Farming was the major occupation (79%) among the referred participants with 41% Anglicans were the majority among the non-referred participants while believers of Bishaka faith were the minority (<5%) in both groups.
Hospital and obstetric factors of women attending Hoima Regional Referral Hospital Table 3. In our study, intern doctors mainly handled caesarean sections for both referred and non-referred mothers attending Hoima Regional Referred Hospital. Emergency caesarean section were the majority in this setting with the most preferred type of incision being midline. Among 61% of non-referred patients, there were no antibiotics given before surgery compared to 42% among referred patients. Premature rupture of membranes was more among referred patients (13%) compared to Premature rupture of membranes among non-referred patients (p=0.03). Participants with more than one previous caesarean section were more in non-referred 29% compared to 25% in referred patient (p = 0.42)  On adjusted analysis: Non-referred mothers who completed primary education were 5.5 times more likely to develop post-caesarean wound sepsis compared to those who never attained formal education (aOR=5.5, 95%CI 1.5-20.2). Non-referred mothers who completed secondary level were 11.6 times more likely to develop post-caesarean wound sepsis compared to those who never attained formal education (aOR=11.6, 95%CI 2.6-51.5). Non-referred mothers who completed tertiary level were 8.7 times more likely to develop post-caesarean wound sepsis compared to those who never attained formal education (aOR=8.7, 95%CI 1.8-43.1).

Bivariate and Multivariate logistic regression analysis of post-caesarean wound
sepsis on associated clinical, hospital and obstetric factors among non-referred patients attending Hoima Regional Referral Hospital Table 6. Mothers who are HIV positive were 8.5 times more likely to develop postcaesarean wound sepsis than those who are HIV negative (cOR=8.5, 95%CI 1.6-44.7).
Mothers who had prior caesarean section were 2.9 time more likely to develop post-caesarean wound sepsis than those without prior caesarean section (cOR=2.9, 95% CI 1.2-  Table 7. Mothers who bathed more than two time were seven times more likely to develop post-caesarean wound sepsis compared to those who bathed two time or less (aOR = 7.0 ,95% CI 1.6-30.2). Referred Bishaka faith mothers were 6.8 times more likely to develop post-caesarean sepsis compared to catholics though it was statistically not significant (cOR=0.6, 95%CI 0.6-71.7). Referred mothers who bathed more than two times were 3.3 times more likely to develop post-caesarean wound sepsis compared to those who bathed two times or less though it was statistically not significant (cOR=3.5, 95%CI 0.9-11.2).
Referred mothers who were underweight were 1.3 times more likely to develop postcaesarean wound sepsis compared to those with normal BMI though it was statistically not  On adjusted analysis: Referred mothers who experienced more than four vaginal examinations were 6 times more likely to develop post-caesarean wound sepsis compared to those in whom four and less vaginal examinations were done (aOR = 6.0 95%CI 1.9-

18.9).
Common bacterial pathogens that were isolated among study participants attending Hoima Regional Referral Hospital Table 9. The most common bacterial isolates were Staphylococcus aureus 28.8%, followed by E. coli, 21%. The least was Proteus mirabilis.

Drug susceptibility pattern of bacterial isolates from septic wounds of mothers in
Hoima Regional Referral Hospital, Uganda  Coliforms showed high susceptibility to imipenem followed by Staphylocccocus aureus to ciprofloxacin.

Prevalence of post caesarean wound sepsis among mothers attending post-natal ward at Hoima Regional Referral Hospital, Western Uganda
In our study the overall prevalence of post-caesarean wound sepsis was 16.8%. This was much higher than the reported Sub-Saharan Africa prevalence of 7.3% [23]. The discrepancy could be due to small area of coverage within the population by the researcher as he only considered Burundi, DRC and Sierra Leone and generalized it to all Sub-Saharan Africa. Though Chu et al. [23] had a bigger sample size compared to this study, this study considered only one referral hospital. In addition, the current prevalence is also lower than that of 22.2% done in Kenyatta National Hospital [24]. This may be due to fact that KNH is the biggest national referral Hospital in East Africa handling more cases than HRRH.
Our study findings show that post-caesarean wound sepsis was higher among referred patients than non-referred patients. This result could be due to complications such as obstructed labour that was higher among the referred patients (48%, 47/98) and this may be attributed to the multiple vaginal examinations before caesarean section and delayed ambulation after surgery. Our findings agree with those by Sai et al where 70% of the cases in Government Hospital, Tirupati, Chitoor district, India were due to obstructed labor [25].

Regional Referral Hospital, Western Uganda
High level of education (primary and post primary) was a significant predictor of postcaesarean wound sepsis among non-referred patients. This is contrary to the findings by Behaviourally, mothers (non-referred) who reported having over 2 baths per day were more likely to experience post-caesarean wound sepsis compared to their counterparts who reported to have two or less number baths per day. This is against biological plausibility and it is possible that baths taken may be unhygienic as our participants were culturally oriented who incorporate herbs in their baths and this increases chances of microbial contamination of the wound post-caesarean section [3]. Additionally, the study participants were from low income background and possibility of having no recommended bathing soap is likely. There is need to sensitise such mothers against use of herbal concoctions in their baths.
Mothers who experienced more than four vaginal examinations had increased risk of post caesarean sepsis compared to mothers who experienced less than 4 vaginal examinations.
Our findings are consistent with findings by Hassan et al obtained among mothers attending Port Harcourt Teaching Hospital, Southern Nigeria [24,29]. The similarity is the highest delivery rate between the two hospitals; Hoima Regional referral Hospital and Port

Common bacterial isolates
We found common bacteria being Staphylococcus aureus followed by E. coli. These findings are in agreement with findings in central Uganda in which Staphylococcus aureus and coliforms have been shown to be common pathogens in post-caesarean wound sepsis [19] . In addition, in Zanzibar, Staphylococcus aureus and E. coli were found to be

Drug susceptibility pattern of bacterial isolates
In our study, coliforms showed high susceptibility to Imipenem followed by Staphylococcus aureus to Ciprofloxacin and this is comparable to a study done in Mbale Regional Referral Hospital, Eastern Uganda [37]. Furthermore, the study also showed that Staphylococcus aureus and other coliforms were resistant to Ciprofloxacin, Ceftriaxone, Gentamycin and Cotrimoxazole. This may be attributed to the fact that these drugs were the most erroneously prescribed medications among the study population in Hoima Regional Referral Hospital. These findings are similar to those found in Greece [39].
The study also showed that Staphylococcus aureus was also resistant to Tetracycline and Penicillin which was in agreement with a study conducted by Bebell et al in a Ugandan referral hospital [40] where over 80% antibacterial resistance to Penicillins among in postpartum mothers was reported. In addition, a study done in Mulago National Referral Hospital revealed a development of Staphylococcus aureus methicillin resistance in postcaesarean wound sepsis [20].

Strengths and Weakness of the study
It is the first documented study done at HRRH to report post-caesarean wound sepsis and associated bacterial organisms. Additionally, people who were re-admitted upon exudate were swabbed, so we minimized loss of follow-up and under reporting of Post-caesarean wound sepsis.
In our study, we did not swab the theater surfaces and aerial space to correlate environmental contamination and post-caesarean wound. We were unable to characterize the bacterial isolates at molecular level due to limited budget.

Conclusions
The prevalence of post-caesarean wound sepsis at Hoima Regional Referral Hospital is 16.8%. The major significant risk factors of post-caesarean wound sepsis were: higher level of education, history of previous caesarean section, hygiene, obstructed labor, HIV seropositivity and multiple vaginal examinations prior to surgery. The most common bacterial pathogens were Staphylococcus aureus and the least was Proteus mirabilis.
Resistance was highest for coliforms and Staphylococcus aureus against, ciprofloxacin, gentamycin, penicillin, ceftriaxone and cotrimoxazole.

Recommendations
The local hospital management should alert the health workers about the high prevalence of post caesarean sepsis at this hospital to devise means of controlling it.
Enhanced awareness to the health workers and the patients on the major risk factors of sepsis and attending to them during management.
Health workers should enhance hygiene amongst mothers before, during and after caesarean section. Community based engagements of Health workers on rational use of antibiotics is extremely necessary.
Additionally, further studies should be carried out to involve swabbing theater and aerial space to correlate environmental contamination and post-caesarean wound sepsis.

Acknowledgement
We appreciate the administration of Hoima Regional Referral Hospital for their cooperation. Appreciation go to Theophilus Pius who assisted in the Microbiology Laboratory. Additionally our sincere appreciation go to Akankunda Sayuni and Mohammed Bituura Agaba for manuscript proof reading and financial support respectively. Finally thanks to Brenda T. Atuheire for providing a conducive room for manuscript writing.

Funding
This project was self-funded through all stages from project design, Questionnaire design, data collection, analysis and manuscript writing.

Availability of Data and Materials
Tables and figures have been provided as supplementary material.
The raw data are available upon reasonable request from the corresponding author.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.    a P-value between prevalence of post-caesarean wound sepsis among referred and nonreferred patients   T a b l e 9: Common bacterial isolates from post-caesarean septic wounds from Mothers in Hoima Regional Referral Hospital, Uganda.  Figure 1 Resistance pattern of bacterial isolates against major antibiotics among mothers attending post-natal ward Hoima Regional Referral Hospital, Western Uganda Susceptibility pattern of bacterial isolates against major antibiotics among mothers attending post-natal ward Hoima Regional Referral Hospital, Western Uganda