In our study, low fresh gas flow anesthesia (≤ 1 L/min) for the neonates undergoing digestive surgery was associated to reduce the incidence of hypothermia. This method could shorten the length of hypothermia, as well as help to alleviate further decreasing of body temperature. Moreover, by multivariable logistic regression performed, we found that low fresh gas flow anesthesia and the length of surgical time were significantly associated with intraoperative hypothermia.
In this study, intraoperative hypothermia was defined as a core temperature (nasopharyngeal temperature) < 36°C, in accordance with the definition of the National guideline in UK [13]. The overall incidence of hypothermia was about 81.9% in our study, which was like the result (83.3%) from a study by Lai [4]. Lai et al. also pointed that the incidence of hypothermia was higher in neonates, compared with infants, toddlers and children [4]. From above evidence, we confirmed that compared to other population, neonates had a faster body heat loss during the intraoperative period due to their physiological characteristics. To our knowledge, hypothermia should be prevented rather than treated after its occurrence. With the increasing awareness of the harms of intraoperative hypothermia, many ways had been proposed to prevent it from happening. In our institution, the strategies for avoiding hypothermia included utilizing anesthesia station with heat and humidity exchanger, warm air circulation, warm infusion fluids, and warm mattress pads and blankets, but the incidence of hypothermia was still high before September 2019 (87.6%), which indicated the available strategies were not enough. To date, some anesthesia providers attributed it to patients’ age, weight and different surgical procedures. Actually, in 2017, Engorn et al. did a retrospective research in neonates and revealed that gestation days, weight, type of surgery, sex, and length of surgery were not the risk factors for hypothermia except the thermoregulation interventions, which pointed that active and passive warming strategies played the pivotal role in maintaining normothermia [12]. A study with multiple logistic regression in Thailand showed that the following risk factors for core hypothermia included high ASA physical status and open surgery. Significant protective factor against core hypothermia was heavier body weight [14]. In our study, the enrolled patients were neonates who underwent digestive surgeries, all of whom had the above risk factors rather than the protective factor, which demonstrated the importance of this study.
Our study proved that the length of hypothermia in low fresh gas flow (≤ 1 L/min) group was less than in routine group and the value of lowest body temperature was higher. As far as we know, inhaled air was warmed and humidified in upper air ways in spontaneous breathing. During mechanical ventilation with tracheal tube, this mechanism was disturbed, and the air was inhaled to lower airway directly without being warmed and humidified. Ventilation with dry and cold gases would lead to a considerable loss of heat from airway. In the Dräger Primus breathing system, to reduce the loss of heat, the exhaled gases moved through the hotplate once before mixing with cold and dry fresh air [15]. In 2011, Castro et al. reported that insertion of a heat and moisture exchanger increased the temperatures of inhaled gas in adults [15]. Moreover, by comparing 1L/min and 3L/min fresh gas flow in a Dräger Primus anesthesia workstation, Bicalho et al. revealed that 1L/min fresh gas flow provided better inhaled gas temperature conservation in children [16], which implied that different gas flow was potentially influence body temperature. However, this theory had not been proven in neonates. In our institution, Dräger Primus anesthesia workstation was used for neonates, but the incidence of hypothermia was still high. Two factors might explain the difficulties to sustain normothermia in neonate even though some active and passive tactics had been applied. First, neonates had a large body face ratio, thin skin, low fat content, which was inconducive to temperature preservation. Of course, this problem was compounded by the effects of anesthetic agents that inhibited central thermoregulation by interfering with these hypothalamic reflex responses. Moreover, non-shivering thermogenesis by metabolism of brown fat was limited in premature or critical neonates who were deficient in fat store. Second, the maintenance of body temperature depended on the balance of heat production and loss. According to our results, we had enough evidence to doubt that compared to other resource, the heat loss from respiratory tract due to high fresh gas flow might occupy a large proportion and 1L/min fresh gas flow might be the critical value. The body temperature was hard to be maintained when the fresh gas flow exceeded 1L/min since heat loss was beyond production, but this was only our suspicion and further studies were needed to verify it.
Previous studies had revealed that hypothermia was correlated with increased perioperative blood loss and rates of surgical wound infection, which might result in prolonged hospitalization and increased incidence of postoperative adverse events [17, 18]. In neonates, the literature about the relationship between hypothermia and prognosis was limited. Our study demonstrated that compared to routine group, the value of lowest temperature was higher, and the duration was less in low flow group, which indicated that low fresh gas flow could alleviate the severity of hypothermia. However, we did not find that fresh gas flow had a relationship with PLOS. The result should be interpreted cautiously because of the shortage of medical resources and insufficient economic development in our region. The guardians might withdraw the treatment because of heavy economic burden, which could artificially shorten length of hospital stay. Besides, other potentially relevant clinical outcomes should be examined in our study—like wound infections, but we could not collect the data about wound infections from our electronic patient registration system. Some surgeons might be reluctant to admit the fact that the wound was infected. Thus, the wound infections were not recorded properly in the chart, especially in the patients with mild wound infections.
Limitations
There were some limitations that should be noted. (1) This was a retrospective study which was at risk of patient selection bias and measurement bias. To reduce the effect of those bias, multivariable regression was conducted to determine the independent association between exposure and outcome. (2) Baseline and preoperative temperature were unknown. When the patients had pre-existed hypothermia, the outcome might be biased. (3) The confidence interval for the "treatment effect" of fresh gas flow was wide, and that a larger sample size would be needed to better delineate the true impact. (4) Because of the shortage of medical resources and insufficient economic development in developing countries like China, most neonates did not have medical-insurance, or the health insurance did not cover newborns. Some critically ill neonates with surgical indications might not be able to receive surgical treatment due to family poverty. Even the patients underwent the surgery, the guardians might withdraw the treatment because of heavy economic burden. In our study, the result of postoperative mortality and PLOS, therefore, might be biased. (5) One potential limitation of the investigation was limited generalizability to other settings.