Acute Kidney Injury and Intraoperative Hypotension in Children: More Questions than Answers

The impact of perioperative hypotension is a hot issue in anesthesia. In adults, intraoperative hypotension is associated with poor outcome, including an increased risk of acute kidney injury. In contrast, in children, there are very few studies exploring any associations between intraoperative hypotension and outcome. In this issue, Schacham et al. report the results from a retrospective cohort study aiming to determine the relationship between intraoperative hypotension and postoperative acute kidney injury in children having noncardiac surgery.¹  They also sought to define any thresholds of hypotension associated with the kidney injury.

Hypotension was defined both in terms of absolute hypotension, the lowest intraoperative blood pressure, recorded for a cumulative period of 5 min; and relative hypotension, assessed as the percentage fall from preoperative baseline, also over a cumulative period of 5 min. Acute kidney injury was defined using standard criteria based on changes in serum creatinine concentrations.²  The cohort study was conducted at two large centers. From a total of 64,412 children who had surgery during the study period, 4,506 were included in the final analysis, of whom 499 (11%) had acute kidney injury. There was no evidence for an association between intraoperative hypotension and postoperative acute kidney injury, and no threshold could be identified.

This result is in some ways consistent with an earlier study by Wingert et al., which did find evidence for an association between intraoperative hypotension and postoperative acute kidney injury in an unadjusted analysis comparing proportions of children that were hypotensive in children with or without kidney injury, but found no evidence for an association between intraoperative hypotension and postoperative acute kidney injury when a multivariable Cox analysis was used.^3.

The most striking aspect of both studies was the high rate of acute kidney injury (11% and 3%). This would seem to go against the often held and perhaps intuitive belief that young kidneys are healthier and more resilient. It must, however, be noted that roughly three quarters of the children had grade 1 acute kidney injury. At least one study has found that this relatively mild degree of acute kidney injury has little if any long-term consequences in children.⁴  This is likely due to most children having more renal reserve than most adults.

There are at least two reasons why the rates seen in these studies may not be representative of the general population of children having surgery. First, urine output was not analyzed in either study. Kidney injury is usually defined by creatinine rise or oliguria. Thus, the rate of injury seen in the study populations may have underestimated the true rate in those study populations. Second, in the study by Wingert et al., only 35% of all noncardiac surgeries performed at the institution over the study period were included in the analysis, while only 7% were included in the study by Schacham et al. These are not representative samples. It would be expected that children who have comorbidity or are scheduled for larger operations are more likely to have preoperative blood draws. Similarly, those with more intraoperative fluid shifts and more complicated operative or postoperative courses will be more likely to have blood draws postoperatively. It is thus likely the general rate of acute kidney injury after surgery in children is much lower than that detected in these studies. Nonetheless, even if all the children excluded didn’t have acute injury, the overall rate would still be around 1%. Given the possible implications of acute kidney injury, this is a worrying number, a worry that cannot be completely dismissed given the paucity of long-term outcome data in these children.

Schacham et al. found that the rate of acute kidney injury was higher in younger children. Once again, this may be at least partly due to bias, as well as the types of surgery more likely performed in each age group. It is plausible that due to the difficulties in obtaining blood from smaller children, clinicians may be less likely to order serum creatinine tests in smaller children, unless the child is thought to be at particular risk. Schacham et al. included relatively few neonates. There is, however, other published evidence that neonates may be particularly susceptible to acute kidney injury after noncardiac surgery, though apart from recognizing the risk, we have only rudimentary data about which surgical neonates are at greatest risk, and few specific evidence-based strategies to prevent the injury.^5–7 

Another limitation of this study is the definition of hypotension. There are various ways to define significant intraoperative hypotension. As well as the actual recorded pressure, we need to consider if it is systolic, mean, or diastolic, whether it is an absolute value or a value relative to a baseline (and defining baseline is itself contentious), the length of time that the child is hypotensive, and whether the hypotensive time is continuous or intermittent. In this study, it was assumed that brief (5-min cumulative) periods of significant hypotension will cause injury. This assumption was based on adult data. The pathophysiology of renal perfusion and injury may be different in children, and it may be that more prolonged periods of lesser hypotension would cause injury. We simply don’t know if this may be the case. One of the drivers for the study was the uncertainty around what is a normal or acceptable blood pressure during anesthesia in children. There are data describing what pediatric anesthesiologists usually do, and what they tend to think, but there is a lack of any evidence for what actually matters.^8–10  Obviously a very low blood pressure will inevitably cause renal ischemia and poor outcome, but unfortunately in this study, the authors did not achieve their goal in determining what we should take as an acceptable level of hypotension in terms of kidney injury. Similar uncertainties surround the level of hypotension that results in significant cerebral ischemia.¹⁰ 

So, what should readers take away from this study? First, although the reported incidence of acute kidney injury may be biased, it does highlight that children are not immune to acute kidney injury, and the incidence may well be higher than we appreciate. Perhaps, as a first step, we need to have a lower threshold for measuring renal function in children postoperatively. Second it might be tempting to conclude that hypotension is not an issue for renal function in children, but I think that may be premature given the overall paucity of data. What the study does highlight is that children are indeed different than adults and that adult paradigms of perioperative care cannot be automatically extrapolated to children. Kiddie kidneys are not little adult kidneys.