CLINICAL PROFILE AND OUTCOME OF CHILDREN WITH SEPSIS IN A TERTIARY CARE CENTRE IN EASTERN NEPAL: A PROSPECTIVE OBSERVATIONAL STUDY

1* 1 1 2 Mukesh Bha a , Shyam Prasad Kafle , Basant Rai , Rejeena Subedi Bha a M et al Received : 20 July, 2021 Accepted : 12 December, 2021 Published : 21 February, 2022 ISSN: 2542-2758 (Print) 2542-2804 (Online) 1595 Birat Journal of Health Sciences Vol.6/No.3/Issue 16/Sep.-Dec., 2021 Original Research Ar cle 1596 ISSN: 2542-2758 (Print) 2542-2804 (Online) Birat Journal of Health Sciences Vol.6/No.3/Issue 16/Sep.-Dec., 202


INTRODUCTION
Sepsis involves a systemic inflammatory response syndrome (SIRS) in presence of infec on, leading to sep c shock and 1,2 mul -organ system dysfunc on. Most deaths caused by 3,4 infec ons are due to sepsis. Sepsis is the most common cause of death in children [5][6][7][8] worldwide. Its mortality rate in children in developing [9][10][11][12] countries is higher than fi y percent. The varied clinical presenta on of sepsis and unavailability of a defini ve test remains a big challenge, hence early recogni on and treatment can be life-saving in developed and pre- 13,14 developed countries. Knowing the clinical profile and outcome of sepsis in children will help in early recogni on, be er management, prognos ca on and mely preven on of complica ons. The study was done with an objec ve of assessing the clinical profile, including the presen ng clinical features, laboratory work up, e ology, treatment, response to treatment and final outcome.

METHODOLOGY
This was a Prospec ve observa onal study carried out at the Pediatric Intensive Care Unit (PICU) of B.P. Koirala Ins tute of Health Sciences (BPKIHS). Data collec on was done over a period of six months from September 2020 to February 2021. The ethical clearance was obtained from the Ins tu onal Review Commi ee (IRC) BPKIHS (Ref no. th 57/077/078-IRC dated on 28 August 2020), and informed wri en consent was obtained from the parents in local language before each enrolment. Pa ents from 1 month to 14 years of age admi ed to the PICU of BPKIHSwith a diagnosis of sepsis by clinical and/or laboratory parameters, and whose parents gave consent were included in the study. However, children with a PICU stay of less than six hours, those with a surgical cause (pre-opera ve or post-opera ve) of sepsis like acute appendici s, blunt trauma abdomen, ulcer perfora on, intes nal obstruc on, trauma, or those who have had inpa ent treatment received in other centre for more than 3 days were excluded from the study. Similarly pa ents who Le Against Medical Advice (LAMA) and those children whose parents refused to give consent were also excluded from the study.

RESULTS
This study was done among 43 children (aged 1 month to 14 years) diagnosed with sepsis. In this study, there were 25 (58.1%) males and 18 (41.9%) females. Thirty one (72%) pa ents were from Terai region, while 12 (28%) were from Hilly region. The median age of the pa ents was three years (IQR 1, 10). The median age among the non-survivors was 2 years (IQR 0.6, 10) while that of the survivors was 4 years (IQR 1, 10.5). Also, the median age between the nonsurvivors and survivors was comparable (p=0.62). The gender distribu on was similar between the non-survivor and survivor group (p=0.89). In this study, 22 (51.2%) pa ents died while 21 (48.8%) survived and got discharged. The most common e ology for sepsis was severe pneumonia, seen in 16 (37%) pa ents, while eight (18.6%) had Gastrointes nal tract infec on, seven (16.3%) Central nervous system infec on, four (9.3%) skin infec on and two (4.6%) had bone and joint infec on. The pa ents presented with different clinical features. Respiratory distress was the most common presen ng clinical feature. The different clinical manifesta ons along with their frequency is given in Table 1.    Table 2. Similarly, 32 (75%) pa ents had hyperlactatemia while 10 (24%) developed respiratory acidosis. On further analysis, the number of pa ents with hyperlactatemia and metabolic acidosis were significantly higher in the non-survivor group as compared to the survivor group. Similarly an abnormal chest X ray (presence of pneumonic consolida on or pleural effusion or features sugges ve of ARDS) was seen in 28 (65%) pa ents. The presence of abnormality on chest X-ray was found to be higher in the non-survivor group as compared to the survivor group. The details of these blood gas and chest x-ray abnormali es are given in table 4. The pa ents were treated with supplemental oxygen, intravenous fluids, ionotropes, intravenous an bio cs and other suppor ve therapies, wherever indicated. Steroids were used in nine (21%) pa ents, while 23 (53%) received blood products in some form. An bio cs in some form (oral or parenteral) were received by 23 pa ents (53.5%) within 3 days prior to the hospital visit. The number of an bio cs used were comparable in both the survivor and nonsurvivor group (p=0.64). Likewise, the use of ionotropes was significantly higher in the non-survivor group (p<0.001). However, the dura on of ionotrope use (whether used for ≤ 48 hours or more than 48 hours) was comparable in the survivor and non-survivor group (p=0.71). It was also observed that use of Mechanical Ven la on (MV) and dura on of mechanical ven la on use for less than or equal to 5 days were significantly higher in the non-survivor group. Ven lator Associated pneumonia (VAP) was seen in eight (18.6%) pa ents. Also, the pa ents who died at a significantly higher rate of complica ons (including Ven lator Associated pneumonia (VAP), deep venous thrombosis, pneumothorax, pressure sore, thrombophlebi s, sep c arthri s, empyema, hepa s and hepa c encephalopathy). The details of the different treatment modali es used is given in table 4.

DISCUSSION
The median age of pa ents in our study was 3 years which [15][16][17][18] was similar to other studies. Study done by Ghimire JJ et 19 al however, showed a higher median age around 7 years. The reason for younger children being more infected can be a ributed to the lower immunity in that age group. There was a slight male preponderance in this study which is [15][16][17][18][19][20][21][22][23] consistent to the findings of other studies. We observed a mortality rate of 51% in our study, similar to 15 the observa ons made by Kaur G et al. However Studies done in developed countries by other authorshad a lower [16][17][18]20,21,23 mortality rate (10-20%). The higher mortality rate in our setup could be due to mul ple factors like delay in hospital arrival, poverty, lack of proper infrastructure and skilled manpower. Nearly a quarter of the pa ents in this study had some form of prior comorbidity, and this was higher in the non-survivor group. Similar are the findings 16,20 given by Pedro Tda C et al and Vila Perez D et al. However, Jaramillo Bustamante JC et al and Humoodi MO et al observed a higher incidence of comorbidity (50-75%) in 18,22 their pa ents. The most common e ology of sepsis in our study was pneumonia, similar to what most of the other authors [18][19][20]22,23 observed.
Dissimilar to this, was the finding of Boeddha NP et al, in which the most common e ology of 21 sepsis was fever without focus. The median dura on of PICU stay (4 days) in this study was lesser than what Ghimire 19,22 JJ et al (6 days) and Humoodi MO et al (8 days) observed. The incidence of AKI was much higher (30%) in our study as compared to the findings of Kaur G et al (2% -3.5%) and 15,16 Pedro Tda C et al (2%). However, the presence of AKI was comparable in both the non-survivor and survivor group in these studies. In contrary, Vila Perez D et al in their study, observed that AKI was significantly higher in the non- 20 survivor group. Similarly, the incidence of Conges ve Cardiac failure (CCF) in this study (2.3%) was comparable to 15,16 the findings of Kaur G et al(1%-4%) and Pedro Tda C et al. Also, the presence of CCF was similar in the survivors and non-survivor group. However, Vila Perez D et al in their study, 20 observed a significant associa on between CCF and mortality. Blood culture posi vity for bacteria was seen in nearly 12% cases in this study; reason for this low yield being indiscriminate an bio c use prior to hospital arrival. In contrary, other studies had a 1 6 -1 8 , 2 1 , 2 2 higher yield (20-51%) of bacteria in blood. Staphylococcus aureus was the most common organism yielded in this study while Vila Perez D et al and Boeddha NP et al reported Neisseria meningi dis as the most common organism while Jaramillo Bustamante JC et al reported gram 18,20,21 nega ve bacilli as the predominant bacteria.
However none of the above men oned studies men oned an associa on of blood culture posi vity with mortality. In this study, ARDS was seen in 20% of the pa ents, which 23 was similar to the findings made by Xiao C et al. However, the incidence of sep c shock was much lower (2.1-48%) in 15,18,24 other studies as compared to this study (79%).
In this study, the incidence of severe sepsis was 65%. However, Pedro Tda C et al observed a higher incidence (89%) of severe sepsis than our study while Jaramillo Bustamante JC et al and Wolfer A et al reported a lower incidence of severe 16,18,24 sepsis in their studies (1.6%-25%). The incidence of DIC in this study was nearly 60%. Dissimilar to this, was the finding of Pedro Tda C et al, who reported a 16 lesser incidence of DIC (1%). Likewise, the presence of MODS in our study (56%) was higher to that observed by  19,20 (64-68%). Likewise the median dura on of mechanical ven la on used in this study (3.5 days) was similar to the 19,23 study done by Ghimire JJ et aland Xiao C et al. Also, the number of pa ents in whom ionotropes were used in this study (79%) were comparable to the findings made by 19 Ghimire JJ et al.

CONCLUSION
This study showed that younger children are vulnerable to develop sepsis. Pneumonia is the most common e ology. Most of the pa ents who develop sepsis presented in sep c shock. The presence of prior comorbidity, severe sepsis, ARDS, MODS, DIC, hyperlactatemia, abnormality on chest Xray and presence of metabolic acidosis were higher in those who died. So, in view of these findings, utmost care and proper management should be ins tuted early in those who develop these features. Likewise, prognos ca on and counselling to the parents can be done based on the above findings.

RECOMMENDATIONS
In view of the above findings, utmost care and proper management should be ins tuted early in those pa ents with sepsis who have underlying comorbidity, severe sepsis, ARDS, MODS, DIC, hyperlactatemia, abnormality on chest Xray and presence of metabolic acidosis. Likewise, prognos ca on and counselling to the parents can be done based on the above findings.

LIMITATIONS OF THE STUDY
This study has certain limita ons. Smaller sample size is one of the limita ons of this study. Also, some pa ents had received treatment prior to arrival at our centre, this could REFERENCES have masked many clinical and laboratory findings. Some pa ents arrived at the emergency with sepsis, while some developed sepsis during the hospital stay. This created some disparity in the hospital course. Similar study, with a larger sample size could be done in the future to get a clearer picture.