Introduction
Cardiopulmonary bypass (CPB) is a high risk and the main treating strategy in patients with coronary artery and valvular heart complications. Cardiac surgery is associated with significant morbidities due to arrhythmias, bleeding, stroke, and neuropsychiatric complications (1).
Temperature management is a significant factor during surgeries, which affects the outcomes of the surgery, thus it should be investigated in patients with coronary artery disease (CAD) that are candidates for CPB to obtain the best results.
Hypothermia has been proposed almost 50 years ago as a beneficial strategy in patients undergoing open heart surgery to reduce the consequent myocardial complications and to increase the myocardial protection during surgery. Hypothermia is known as an important approach which decreases cerebral oxygen demand and maintains blood flow in the central nervous system (CNS) during cerebral ischemia occurring in cardiac surgery (2).
Although CPB is conducted based on hypothermia with considerable successful results, various systemic adverse effects are reported to be associated with this surgical temperature including compromised drug metabolism, diaphragmatic dysfunction, long recovery time after anesthesia, cardiac complications, coagulopathy, wound infections, and postoperative shivering  (3). Therefore, normothermia might be able to decrease the incidence of mentioned risks; however this method also elevates the possibility of cerebral and neurological complications. The exact advantages and disadvantages of each method are uncertain and under the investigation.

Methods
Literature search
PubMed was searched for the relevant articles with the following search term: hypotherm* AND normotherm* AND coronary AND (death OR mortality). Irrelevant articles were excluded after studying titles and abstracts following the initial search. The full text of the remaining articles was then studied. Reference lists of the included articles were searched to prevent missing any relevant article.

Study selection
Inclusion criteria were any randomized clinical trial which compared hypothermia with normothermia in patients undergoing CPB, induced cardioplagia, and provided data about the in-hospital death of patients after the surgery. Only English language articles were included in this systematic review. Relevant articles were extracted with no time limitation. Exclusion criteria were retrospective cohort and prospective studies, and case reports. Articles which induced mild hypothermia and the rewarmed patients were also excluded from the results.

Data extraction
We excluded the patients who were died after the cardiac surgery in hypothermic and normothermic groups. In-hospital death or perioperative death were considered as mortality event in each group. Relative risk reduction (RRR), absolute risk reduction (ARR), and number needed to treat (NNT) were calculated according to the number of death in each group.

Results
Overall, 225 articles were identified through the initial search in PubMed; 196 articles were excluded from the results after studying titles and abstracts of the articles. Finally, 29 articles remained as the most relevant articles with the purpose of this study. One article was also excluded due to inaccessibility to full text for retrieving the event rate. The process of inclusion and exclusion pattern of the articles is revealed in Figure 1. Quality assessment of the results is provided in Table 1.
Only one out of 28 included studies specifically investigated the mortality following coronary bypass with hypothermic condition compared with normothermic as its primary outcome. In other studies, mortality rate was reported as the secondary outcome.
The mortality rate was zero in hypothermic and normotehrmic groups of 8/28 included studies, thus RRR, ARR, and NNT could not be calculated in these studies. The calculated RRR, ARR, and NNT for the remaining 20 articles are provided in Table 2. There were no significant differences between investigated groups of each included study regarding the patients’ age, gender, and preoperative conditions.
Discussion
Compared with normothermic temperature, hypothermia is considered to be beneficial in reducing the possibility of brain and other tissue injuries and ischemia. The incidence rate of ischemia and the following permanent damages might be higher during normothermic temperature surgeries (26).
For estimating the risk and advantages of temperature adjustment for CPB, it is essential to consider and compare various clinical variables. Clinical variables mostly studied in articles were cerebral, myocardial, and renal complications, in-hospital stay duration, blood transfusion requirement, and bleeding.
The mortality rate of patient candidate for CPB under the normothermic condition compare with hypothermic, has not been studied before. According to experimental studies, hypothermia has a definite cardioprotective effect, but the results of clinical studies indicate otherwise.
Based on results obtained in one previous systematic review and meta-analysis performed by Rees et al. in 2001, there was a statistically non-significant difference between hypothermic and normothermic methods regarding the perioperative mortality rate; however the hypothermic approach was associated with higher number of non-stroke perioperative death. The estimated Odds ratio (OR) was 1.46 [0.90, 2.37] in hypothermic versus normothermic methods. Data in that systematic review was extracted from 16 studies investigated total of 4201 patients and the revealed information was about perioperative deaths not caused by stroke (30). In that systematic review the follow-up duration was within 30 days after the surgery.
Only 1 out of 28 included trials, specifically studied the mortality rate following hypothermia condition compared with normothermia in patients undergoing aorto-coronary bypass surgery in a large sample size clinical trial conducted in three center in 1994. In this study, randomization was performed by envelope method. The follow-up duration was 30 days (26).
Mortality rate is an important factor that could change the treating approach under CPB surgery. According to articles, in-hospital mortality is defined as death during the hospital stay. Based on included studies nasopharyngeal and rectal temperature probe was the most reliable indicator of brain temperature.
Based on included studies in present systematic review, death occurred due to various reasons including cardiogenic shock following perioperative myocardial infarction at the day of surgery, hyposensitivity and not being resuscitated, low cardiac output and cerebral infarction, sudden death on postoperative day. In one study in 1994, patients death were screened in two different categories including deaths caused by myocardial infarction or ischemia (refractory cognitive hearth failure without postoperative myocardial infarction), and deaths caused by other reasons related to ischemia (sepsis, multiorgan failure following prolonged ventilation) or not related to cardiac events (24).
Another article studied the effect of postoperative hypothermia on patients underwent CABG regarding incidence of death. In this study, in patients with intensive care unit (ICU) admission, temperature lower than 36°C considerably had a higher mortality rate compared with normothermic patients; odds ratio, 0.59; 95% confidence interval, 0.39-0.91). Therefore, they concluded that patients who arrived at ICU with normal temperature were at lower risk of in-hospital mortality. Totally, their results showed that there was not a considerable difference between hypothermia and normothermia considering the mortality rate. In this regard, mortality rate of these approaches is not high enough to change the chosen treating strategy in patients who are candidates for CABG. This non-significant difference between two studied approaches might be due to low sample size of the included studies or due to the treating strategies performed routinely in ICU following cardiac surgeries which reduce the death incidence.

Conclusion
Due to the small number of death among screened articles, clinical trials with larger sample sizes might be able to reveal the mortality rate difference between two surgical temperatures.

Acknowledgement
We would like to thank Clinical Research Development Unit of Ghaem Hospital for their assistant in this manuscript.

Conflict of Interest
The authors declare no conflict of interest.