Surgical Apgar Score, Predictor of Post-Emergency Abdominal Surgery Outcome

To provide consistently low incidence of major complications for patient undergoing any operation, recognizing patients at high risk of developing a complication will contribute substantially to quality of operation and of cost reduction in surgery. Marked variability of post-operative outcomes is usually found due to difference in patient’s preoperative co-morbid and risk factors [1]. The cardinal objective in the emergency situation is preparation and resuscitation for correction of deranged physiology. Inadequacy of which is a major contributory factor to morbidity and mortality. Intra-operative variable such as heart rate, blood pressure, arterial saturation, tissue perfusion and the surgical stress further contribute to variability in patient’s risk of developing complications. However, the degree to which intraoperative performance further contributes to variation in patients’ risk of complications remains unclear [1]. In operating room, surgeons have relied principally on "gutfeeling", instead of their objective assessment of the operative course for postoperative prognostication [2]. The Surgical Apgar Score (SAS) was determined by [3]. The SAS has been validated mainly in the west but less numbers of studies are currently available from Hospitals in our country. Here we prospectively evaluate postoperative complications according to SAS score.


Introduction
To provide consistently low incidence of major complications for patient undergoing any operation, recognizing patients at high risk of developing a complication will contribute substantially to quality of operation and of cost reduction in surgery. Marked variability of post-operative outcomes is usually found due to difference in patient's preoperative co-morbid and risk factors [1]. The cardinal objective in the emergency situation is preparation and resuscitation for correction of deranged physiology. Inadequacy of which is a major contributory factor to morbidity and mortality. Intra-operative variable such as heart rate, blood pressure, arterial saturation, tissue perfusion and the surgical stress further contribute to variability in patient's risk of developing complications. However, the degree to which intraoperative performance further contributes to variation in patients' risk of complications remains unclear [1]. In operating room, surgeons have relied principally on "gut-feeling", instead of their objective assessment of the operative course for postoperative prognostication [2]. The Surgical Apgar Score (SAS) was determined by [3]. The SAS has been validated mainly in the west but less numbers of studies are currently available from Hospitals in our country. Here we prospectively evaluate postoperative complications according to SAS score.

Materials and Methods
This study was conducted at Department of General Surgery in Mamata General Hospital, Khammam, Telengana, India, a tertiary care teaching institute and hospital. The target population was patients undergoing laparotomy admitted to the general surgical wards, who met the eligibility criteria. This was a hospital based, single centre longitudinal prospective observational study from October 2014 to March 2015 including 66 patients.
Inclusion criteria: All patients above 16 yrs undergoing emergency abdominal surgeries under general anesthesia and willing to participate in study.

Exclusion criteria:
Surgeries under spinal / epidural / local anesthesia, not requiring intensive monitoring and regular follow-ups.

ASA (American Society of Anaesthesiologists) Class.
Surgical procedure executed.
Intra-operative variables for surgical apgar scoring (Table I).
Major complications or deaths within 30 days of surgery (End point of study).
The SAS was calculated at the end of emergency open abdominal surgery from the EBL, lowest MAP and lowest H.R. recorded during the operation. The score was the sum of the points from each category. Data such as lowest heart rate and lowest mean arterial pressures reached during the procedure were collected from the anesthesiologist's records (manual). Occurrence of pathologic bradyarrhythmia, including sinus arrest, atrioventricular block or dissociation, junctional or ventricular escapes rhythms, and asystole, also receives 0 points for lowest heart rate. Extra physiologic values of HR (<20/min or >200/min) and MAP (<25 mmHg or >180 mmHg were excluded. Estimated blood loss was calculated using the formulae [5]. Blood loss=[ (EBV x (Hᵢ -H f ) / {(Hctᵢ + Hct f ) /2} ] + (500xTᵤ) [5] [Estimated blood volume (EBV) is assumed to be 70 cm 3 /kg, Hᵢ and H f represent pre-and post-operative haemoglobin, Hctᵢ and Hct f represent pre-and post-operative hematocrit, Tᵤ is the sum of autologous whole blood (ABW), packed red blood cells (PRBC), and cell saver (CS) units (FFP, Cryoprecipiatate) transfused].
Surgeries performed were emergency exploratory laparotomies (adhesiolysis, intestinal perforation repairs, resection and anastomosis for gangrenous bowel, appendectomies for acute appendicitis (perforated with or without gangrenous changes), Hernioraphies for obstructed ventral hernias, splenectomy for grade 3, 4, 5 splenic injuries due to blunt abdominal trauma. Patients were subsequently grouped into three categories based on their SAS for purposes of risk stratification, as in the study by Michael Dullo [6] which was as in the following table (Table Ia).

Data management and analysis
All data were expressed as mean ± SD (Standard Deviation). Student's t-test was used for comparing between the two groups. The statistical significance between all the groups was evaluated by Oneway ANOVA and with Bonforroni's post-hoc test using SPSS Software Version 16.0.

Ethical considerations
Ethics and Research Committee reviewed the study protocol and granted approval prior to commencement of study.

Result
Demographic profile has been shown below (Table II and Table III).
In table II the trend of mean SAS increased as the age range decreased from more than 60 years to less than 40 years but it was not statistically significant (P=0.84) (Table II).
In table III mean SAS was not significantly related to gender (P=0.563) ( Table III).
As mentioned in the table IV most common co-morbidities and risk factors were smoking (36.4%), hypertension (31.8%) and diabetes mellitus (28.8%). Hypertension, diabetes mellitus, renal failure were highly significant (p<0.001) for major complications (Table IV).
As mentioned in the table V the rate of occurrence of major complications increased as the number of preoperative co-morbidities increased among patients. Patients with 5 co-morbidities had 5-fold increases in complication rate compared to patients with single co-morbidity (100% vs 20%). As the number of preoperative co-morbidities increased among patients, mean SAS trends to be of lower value, which was proven highly significant (P<0.0001). Patients with 4 or 5 co-morbidities had mean SAS <4, whereas patients with nil preoperative co-morbidities had mean SAS >6 (Table V).
The study included emergency laparotomies with 3 types of intraoperative pathological findings viz. acute intestinal obstruction, hollow viscous perforation, blunt trauma abdomen causing splenic injury, of these acute intestinal obstructions with/without gangrene was most common (31.8%) followed by gastro-duodenal perforation (18%). Major complication including mortality was highest in hollow viscous perforation with peritonitis (21.2%) followed by acute intestinal obstruction (13.6%) and splenic trauma (9.1%).
As mentioned in the table VI mean LHR and mean EBL were significantly related to SAS risk groups (p<0.0001), which meant a patient with low mean LHR and low mean EBL were associated with low risk (Table VI).
Most common major complication were pulmonary diseases (34.8%) including pneumonia (19.7%), prolonged ventilation (12.1%), pulmonary embolism (3%), followed by deep vein thrombosis, acute renal failure, wound disruption, sepsis and shock. Table VII showed, there was significant association between type of risk and occurrence of complication (P<0.001). The cell of high risk group with complications had standardized residuals greater than 1.96. So, this cell contributed the most in predicting occurrence of complications i.e., high risk group patients were the strongest predictor of developing complications. Mortality was considered Clavien grade V of major complications. Out of total 7 deaths, 71.4% belonged to high risk group 28.6% patients in medium risk group (Table VII).
Among the patients in high risk group, 88% had multiple complications and one patient had maximum 5 complications. Death     occurred in 27% of patients in high risk group; most common cause was sepsis secondary to peritonitis on average postoperative day (POD) 5.5. Among the patients in medium risk group, 5% had multiple complications and one patient had maximum 2 complications. Death occurred in 5% of patients in medium risk group; most common cause was sepsis secondary to peritonitis and pneumonia; on average POD 7. Among low risk group no patient developed postoperative major complication.
As mentioned in the table VIII mean SAS for the groups of patients with post-operative complications in decreasing order showed gradual increase in trend respectively, which was statistically significant (P<0.0001). Thus, it was proven that, there was highly significant correlation between mean SAS and outcome in terms of development of major complications including mortality. So, patient more prone to develop major complications including mortality in 30 day follow up had low mean SAS (high risk). Mean ASA was found to be inversely proportional to SAS, which gradually increased as the SAS decreased. The published absolute mortality rates of the individual classes of ASA shows considerable variation, and with increasing mortality rate as classes number increases from I to V [7]. This correlates reverse association with SAS (Table VIII).

Discussion
The calculated surgical apgar scores ranged from one to nine and n-no. of patients, * p<0.05(significant), ** p<0.01(moderately significant), *** p<0.001(highly significant), NS-non-significant     had an average of 5.3 in this study. This was similar to studies by [6] in which SAS ranged from one to nine with a mean of 5.0. In studies by [8] surgical Apgar scores ranged from 6 to 10 with average of 8. Table IX showed, the distribution of patients among different risk groups in our study were as per other literatures (Table IX).
From demographic data, it was found with decreasing scores, patients were increasingly older and more likely to be male which was comparable with studies by [11] but it was not statistically significant in present study. In study by [8] all demographic data were unrelated to SAS. Table X showed the distribution of intra-operative parameters of patients among different risk groups in our study were as per other literatures. Mean EBL, mean LMAP and mean LHR in the group of perforation peritonitis with major complication were 697.2 ± 488.2, 76.2 ± 13.1 and 69.1 ± 12.8 respectively. Similarly, mean EBL, mean LMAP and mean LHR in the group of splenic trauma with major complication were 1688.4 ± 868.9, 61 ± 12.4 and 89.2 ± 18.8 respectively and mean EBL, mean LMAP and mean LHR in the group of acute intestinal obstruction were 360.4 ± 304.4, 72.3 ± 10.1 and 78.1 ± 10.7 respectively. The reason of major loss of blood among splenic trauma cases was higher grade of injury like grade IV and V among the patients who underwent splenectomy. In present study, among some patient before proper resuscitations blood sample was sent for investigation, thus the value of haematocrit reflected was from concentrated blood sample. The haematocrit value after surgery was in fact true value because of intraoperative proper resuscitations. Damage control resuscitation as commonly part of management of blunt trauma cases, where patient was shifted to operation theatre even before complete resuscitation as happened among blunt trauma abdomen with splenic injury in present study where already patient had lost enough blood in intra-peritoneal compartment. As the EBL depended up on pre-and post-operative haematocrit, so above 2 elements already discussed most likely gave fallacy for estimated result. We found discrepancy of blood loss between present study and other literatures, as present study included only emergency surgeries ↓ GA (Table X).
In the table XI rate of occurrence of major complications was higher in present study (94.5%) compared to other studies [3,6,11,12] this could probably be explained as present study included only emergency cases done under general anaesthesia with no prior preparation for comorbid conditions and risk factors, where as other studies included both emergency and elective cases (Table XI).
It was demonstrated about the significant association (P<0.001) between type of risk and occurrence of complication as per table VII and the ability of the SAS in identifying patients at higher than average risk of major post-operative complications as per table VIII. Thus, the predictive value of surgical apgar score to predict the complications was found to be significant. Patients with low scores were more likely to suffer multiple complications (p<0.0001) as in study by [11]. In present study 56% patients with nil post-operative major complications, 15% patients with single major complication, 29% patients with multiple major complication had mean SAS of 6.3, 5.2 and 2.5 respectively.  (Table XII).

Conclusion
Thus, it was proven that SAS was significant predictor of outcome in terms of postoperative major complications including mortality in 30-day post emergency surgeries under general anaesthesia.