Statistics and SIDS

I have just finished reading Statistics: A Very Short Introduction by David J. Hand.  Overall, there wasn’t much practical information to work with, but it did provide a very basic background in general statistical theory, Bayesian networks, and hidden Markov models.  Throughout the book, Hand emphasizes the success of modern computing in the field, gives practical uses for each concept introduced, and explores a range of real-world applications of statistics.

One of the real-world instances chosen by Hand is the probability a mother could lose two of her children to SIDS, aka crib/cot death.  In 1996, UK resident Sally Clark had lost her first son to SIDS.  When her second son passed away in the same manner in 1998, she was arrested and tried for murder. Pediatrician Sir Roy Meadow testified that the chance of two children from an affluent family suffering sudden infant death syndrome was 1 in 73 million. The reasoning for such a low probability was that the second instance of crib death was considered to be independent of the first. If the two incidents are independent, the probability of one instance (1 in 8500) can be multiplied by itself to yield Meadow’s figure. However, as an article written in Plus Magazine explains, it is more reasonable to calculate the conditional probability using Bayes theorem to determine the probability of her innocence.  Which they conclude to be 2/3, given number of births and double murders each year in England and Wales.  Unfortunately, it wasn’t until the second appeal that Sally Clark was exonerated of her charges and released from prison in 2003 after serving 3 years of her sentence.

It was with great frustration that mere hours after reading about Sally Clark’s ordeal that I heard a similar story play out on an NPR broadcast.  In 2008, 17-year-old Nga Truong was interrogated and coerced into confessing to the murder of her own child.  This time is was her brother that had passed alway when she was 8 due to SIDS.  Under the same presumption that two incidents of SIDS in one family is improbable, the detectives aggressively interrogated Nga Troung while accusing her of both deaths.  Troung had been locked up for 2 years before her confession was suppressed on the basis of coercion.  The 13 minute story and police recording is available through NPR.

Taking in both these incidents in the same day was very troubling, primarily because the police in Troung’s case were more determined to get a confession than get the truth.  Even more troubling is the fact that Sally Clark died of alcohol poisoning in 2007, having never recovered from the incident.  While people like Donna Anthony, Angela Cannings, and Trupti Patel were all found innocent after further review of their cases, one has to wonder how many mothers have to face this turmoil without proper defense.