Patient Outcomes

Over the past decade 98 children with short bowel syndrome treated at Mott Children's Hospital were analyzed for their outcomes and predictors of survival. Of these children, 56% were secondary to congenital disorders, 35 were acquired as neonates, and 9% developed SBS later in life. Most patients (63%) weaned off TPN, and 22% died, and the remainder are still on TPN with or without some enteral intake.

Risk of Death

Relative risk (RR) of death was strongly correlated with level of highest conjugated bilirubin (RR=1.1, confidence interval (c.i.) =1.07-1.014,P<0.0001). Interestingly, the small bowel length, number of septic events, cause of SBS and presence of an ileo-cecal valve (ICV) were not significant.

Ability to Wean Off TPN

The ability to wean off TPN was significantly correlated with small bowel length:

(RR=1.004, c.i.=1.000-1.008,P<0.05)
ICV presence (RR=0.52,c.i.=0.29-0.91,P<0.02)
and number of septic events (RR=0.94,c.i.=0.89-1.00,P<0.04)

Cholestasis

Further analysis to determine factors contributing to cholestasis showed that only small bowel length correlated (RR8.8, c.i.6.2-11.6, P<0.01). Gestational age, birth weight, number of septic episodes and etiology of short bowel syndrome did not correlate with the development of cholestasis. 10 children underwent a bowel lengthening procedure with improved outcome in 8. Eight were referred for bowel transplant with only two achieving a successful outcome.

Clearly, despite advances in SBS care, overall failure remains high. Future therapies which can prevent sepsis and cholestasis will be essential to improving these outcomes.

The graph shown below displays a Cox regression analysis of survival after stratifying patients based on whether they had a conjugated bilirubin greater or less than 2.0 mg/dL.

Figure 15

Other Prognostic Factors

The following tables show factors which were examined for contributing to pediatric SBS survival. As can be noted, only small bowel length, at time of diagnosis, was predictive.

Table 4 Covariate Risk Factors for Survival of Short Bowel Syndrome Children

Covariate	Beta Coefficient	95% Confidence Intervals	P Values
Cholestasis	-3.492	0.004, 45.239	0.0001
Number of Septic Episodes	0.998	0.890, 1.117	0.966
Ileo-cecal Valve	4.194	0.897, 19.624	0.069
Small Bowel Length	0.006	1.002, 1.010	0.007
Diagnosis	0.298	0.488, 3.72	0.564

Factors associated with weaning from TPN

The Table and Figures below show the analysis of factors which have been correlated to being able to be weaned from TPN.

Table 5 Covariate Risk Factors for Predicting Ability to Wean off TPN

Covariate¹	Beta-coefficient	95% Confidence Intervals	P Values
Cholestasis	1.022	0.982, 1.063	0.280
Number of Septic Episodes	0.949	0.896, 1.005	0.072
Ileo-cecal Valve	0.534	0.302, 0.944	0.031
Small Bowel Length²	0.557	0.321, 0.968	0.038

¹Defined as < 70cm vs. > 70cm.
²Other parameters, not shown, and were not significant, included etiology of SBS, birth weight and gestational age.

Figure 16 below is a Cox Regression of patients weaning off TPN based on length of small bowel. Bowel length is divided between infants having < 70 cm vs. > 70 cm of bowel length at time of diagnosis. Bowel length > 70cm was significantly associated with the ability to wean off of TPN (P =0.038).

Figure 16

Figure 17 below is Cox Regression of patients weaning off TPN based on presence of absence of ileo-cecal valve (ICV). Presence of an ICV was significantly associated with the ability to wean off of TPN (P=0.031).

Figure 17