Risk Factors Associated with in Hospital Complication Post Gastrointestinal , Pancreatic , Hepatic Cancer Surgery : A Retrospective Case Control Study ( RAPHA Study )

Background : This study aims todevelop a scoring system that will predict in-hospital morbidity post gastrointestinal (GI), pancreatic, hepatic cancer surgery in adult patients. This study took place in De La Salle University Medical Center. Method: Two hundred eighty five (285) adult patients 18 years old and above who underwent gastrointestinal, pancreatic and hepatic cancer surgery from 2010 to July 31, 2014 were included. Variables were evaluated in the univariate and multivariate analysis. Calculation of specific score from the resulting factors was performed by logistic regression analysis to develop the scoring system and to determine the best cut-off score in predicting in-hospital morbidity. Results: Out of 142 patients with post-operative complications, factors significantly associated with morbidity were as follows: age ≥ 75 years (p = 0.002), low serum albumin (p = 0.00), abnormal electrocardiogram (ECG) findings (p = 0.036) and emergency surgery (p = 0.000). Calculated best cut-off score was 1.4. Conclusion: The RAPHA scoring system may serve as a promising aid in predicting morbidity and mortality among patients who will undergo GI cancer surgery.

There are several established scoring systems which are currently being utilized to assess patients for general surgery pre-operatively.Collaborative efforts from internists, surgeons and anesthesiologists has aided creating such scoring systems to assure patients safety and measure the probability of complications. 8sk Index were some of the pre-operative scoring Bollschweiler et al evaluated the use of physiological and operative severity score for the enumeration of mortality and morbidity (POSSUM) scoring system in patients undergoing D2 gastrectomy among gastric cancer patients. 9This scoring system has been created by Copeland et al in 1990 as patient risk prediction model constructed on 12 patient characteristics and 6 operative characteritics.It was developed for quality assessment in general surgical units.There were surgeries (V-POSSUM, CR-POSSUM). 10 Its use has also been validated to assess risks with other surgical procedures such as orthopaedic, pancreatic and gastric surgery.This scoring system includes intra-operative variables and therefore cannot completely assess patient pre-operatively.Moreover, there is no scoring system post GI, pancreas, liver cancer surgery pre-operatively.
pre-operative risk assessment on patients who will undergo gastrointestinal malignancy surgery.

INTRODUCTION
Cancer is the third leading cause of morbidity and mortality in the Philippines.Among the top listed cancers are cancers of the oral cavity, stomach, colon, liver, and rectum.In year 1993-1995, the Department of Health-Rizal Cancer Registry (DOH-RCR) and the Philippine Cancer Society Inc.-Manila Cancer Registry (PCSI-MCR) presented that liver malignancies were more predominant with 40 individuals per 100,000 population diagnosed. 1Worldwide, gastrointestinal (GI) malignancies remains a public health issue.Virani et al discussed the prevalence of GI tract and liver cancer. 2Colorectal cancer (CRC) has an annual incidence of 1 million cases and an annual mortality of more than 500,000 cases.CRC is the second most common cause of cancer death followed by gastric cancer.Cancer of the pancreas is the fourth most in women.It has the lowest 5-year survival rate of any gastrointestinal tumors.Hepatocellular carcinoma (HCC) is the sixth most common cancer in the world and the third most common cause of cancer mortality. 3ccompanying the number of diagnosed cases procedures done in the surgical world.All the demographics and variables data were collected retrospectively from each patient's medical record.Primary outcome and the cases would be those who will develop in-hospital morbidity post GI, pancreatic, hepatic surgery.Controls would be those without morbidities while being admitted in the hospital postoperatively.In-hospital mortality among the groups will also be computed.were entered simultaneously into multiple logistic regression models.Independent prognosis factors emerging from these logistic regression models were used to develop the scoring system to determine the best cut-off score in predicting in-hospital morbidity.
Best cut-off score was determined using an ROC at each cut-off point.
All of the variables in this study were the standard usual data in the chart.As this is a retrospective study, all information were from data written in the medical record.Only complete medical record were included in this study.Although there was a possibility of missing data bias, this study was still important as a basis to morbidity and mortality post operatively in GI surgery patient.In order to anticipate missing data bias, all incomplete medical records were recorded separately and analyzed.The scoring system developed in this study may be externally validated with other further retrospective or prospective study.

RESULTS
A total of 285 patients were included in the study where in 143 patients had no morbidity, while 142 were considered under the group of patients who developed complications post-operatively.Moreover, among those who had morbidity, 19% of them had severe complications and eventually died.Among the 81% who were alive, most of the noted complications were acute kidney injury and hospital acquired pneumonia, which comprised 49.6% and 22.6% respectively.Univariate analysis was performed and results showed that only some of the variables below affected results showed that age (p = 0.001), hemoglobin (p = 0.060), WBC (p = 0.001), creatinine (p = 0.007), serum sodium (p = 0.018), serum potassium (p = Specifically, results showed that those who were elderly (75 years old and above) had higher chances of morbidity.Similarly, those with abnormal values of Hemoglobin, WBC, serum creatinine, serum sodium, serum potassium and serum albumin were more prone to develop complications post-operatively.Furthermore, those who had ECG findings other than normal sinus rhythm were more likely to have morbidity as well.Among the subjects who developed complications post-operatively, most common ECG findings noted were ischemic changes as well as rhythm abnormalities (Table 3).Similarly, those who underwent emergency surgery had higher chances of having complications.
After considering all the risk factors that were morbidity.The results showed that those whose age was at least 75 years old was  The resulting p of 0.0001 suggest that the area under 0.50 area.This denotes that the new scoring system of morbidity.Area under the curve = 0.826, p = 0.0001 1.4 is expected to have morbidity.It has a sensitivity

DISCUSSION
This study showed that in patients who underwent GI, hepatic, pancreatic cancer surgery, age 75 years and older, abnormal albumin levels, abnormal ECG higher risk for complications.
As shown in previous studies, age itself is an independent risk factor in the development of complications post operatively. 13In this study, age more than or equal to 75 years is at higher risk for morbidity representing 21% of all patients who developed complications.the post-operative status of GI cancer patients.Low albumin levels were associated with poor wound healing, decreased collagen synthesis and impairment of immune response.All of these events lead to infections such as surgical site infections, pneumonia and anastomotic leakage. 14In this study, abnormal values of serum albumin is also a strong predictor of morbidity among patients who underwent post GI cancer surgery (p = 0.00).Seventy percent (70%) of patients who developed complications post-operatively had lower albumin values.This study also showed that majority of the morbidities noted were due to hypoalbuminemia.These morbidities included hospital acquired pneumonia, intraabdominal infection and surgical site infections.
Abnormal ECG findings were associated with post-operative complications as stated in the study by Igari et al. 14 Similar results were noted in this study.The ROC curves suggested different cut-off score.The table below reveals that the best cut-off score which gives the highest combined score of sensitivity Almost 53% of patients with abrnormal ECG results developed complications after the surgery.Most of the and ischemic changes.Forty nine percent of the patients who died due to complications had abnormal Emergency surgery, as mentioned earlier, plays an important role in determining postoperative mortality and survival among cancer patients who undergo surgery.In a study by AL-Homoud et al, post operative mortality reaches about as high as 12% (95% CI: 10.6-13.4). 15redicts a higher chance of developing complications in post GI cancer patients (p = 0.000; 95% CI: 2.84-23.19).
For the RAPHA scoring system, ROC curve was plotted and it showed that the best cut-off score was 1.4.years old, abnormal albumin levels, abnormal ECG develop complications post gastrointestinal, pancreatic and hepatic cancer surgery.
All data in this study was gathered retrospectively from the medical records, which is prone to recall bias and missing data bias.Another limitation of the study is that it was performed in a single center.A multicenter study is recommended for better representation of the population.Furthermore, a prospective study with a larger study population is recommended for obtaining better measure of association for some variables and validation of the scoring system.

CONCLUSION
The RAPHA scoring is a promising aid in predicting morbidity and mortality among patients who will undergo GI cancer surgery.It is important to note that this scoring system was developed to assist the internists, surgeons and anesthesiologists in evaluating the risk of these patients for complications.In a way, this tool might somehow prevent such morbidities to happen or even to lessen the rates of these complications to develop post operatively.

Table 3 . Subjects groups
Risk Factors Associated with in Hospital Complication Post Gastrointestinal, Pancreatic, Hepatic Cancer Surgery: A Retrospective Case Control Study (RAPHA Study)