Pulse oximetry screening for critical congenital heart disease: a comparative study of cohorts over 11 years

Background: Pulse oximetry (POS) has been proposed as a screening tool for CCHD in newborn. The aim was to identify the effect of POS on the rate of diagnosis after discharge and survival to one year in cases with CCHD. Material and Methods: All cases of CCHD from three tertiary level hospitals in the Northern region of UK between 1st January 2001 and 31st December 2011 were identified from the Northern Congenital Abnormality Survey (NorCAS). A retrospective cohort study comparing screened and unscreened population for CCHD was undertaken. The main outcome was post discharge diagnosis rate and mortality at one year between the cohorts. Results: Total number of births during the 11 years was 138,176. A total of 147 cases had CCHD, 59 diagnosed postnatally. Five and eight cases were diagnosed after discharge in the screened and the unscreened cohort respectively. The rate of post-discharge diagnosis in the screened population was 7/100,000 and 13/100,000 in the unscreened population with a relative risk of 0.52 (CI 0.2 to 1.42). Mortality at one year in postnatally diagnosed cases was five and one in the screened and unscreened cohorts respectively. Conclusion: With good antenatal detection rates, POS did not have a statistically significant impact in identifying cases of CCHD, when added to the present screening process of antenatal ultrasound and postnatal examination. The same conclusion cannot be made for regions with lower antenatal detection rates; perhaps it may be more appropriate to consider pulse oximetry as a screening tool for hypoxemia of any cause. Brief rationale This is the first study evaluating the contemporaneous post-discharge diagnosis rate between screened and unscreened populations. The rate of post-discharge diagnosis was 7/100,000 in the screened and 13/100,000 in the unscreened populations. However, this did not achieve statistical significance and in a setting with high antenatal diagnosis a very large study would be required to demonstrate efficacy of POS.