Life-saving claims need better quantification

The National Treatment Agency for Substance Misuse (NTA) has today published a report saying that 18 lives have been saved as a result of a programme in which the friends and family of addicts were given training in the use of the heroin antidote naloxone. By training those likely to be near at hand when an overdose occurs, hundreds of lives could potentially be saved in the UK every year, said the agency’s chief executive Paul Hayes.

The report is largely qualitative, dealing with the problems involved in setting up the training programme and how they were resolved. But better quantification would also have been good.

Between July 2009 and February 2010, the agency funded 16 pilot sites around England to train family or carers on overdose recognition, intervention, and how to administer the intramuscular injection of naloxone.

Most sites took a couple of months to get established and one issued no naloxone. Participants were asked to complete a baseline A-questionnaire and a post-training B-questionnaire.  At the 16 sites, 495 carers were trained to respond to overdoses. The report by the NTA does not specify:

  1. how many Naloxone kits were issued to carers
  2. how many A-questionnaires were completed
  3. how many B-questionnaires were completed
  4. how many trainees completed both their A & B questionnaire
  5. how the NTA learned about the reported 18 administrations of naloxone
  6. how many of the 18 administrations were to the prescribee
  7. how many of the 18 administrations were by trained carers
  8. whether all of the 18 administrations had occurred by end February 2010
  9. the cost of the 16 pilot studies.

We may estimate the rate of naloxone administrations per trainee-month as follows. Sixteen sites trained 495 carers in 6 months: just over 5 trainees per site per month.

All but one of the sites (15) issued naloxone. Assume that at most 464 naloxone kits were issued to trainees by consent of the patient for whom naloxone was prescribed.

Assume that the 18 administrations of naloxone that the NTA counted were all by trained-carers and all had occurred by February 2010. Assume a mean of 3.5 naloxone-administration months per carer-trainee and so 464 X 3.5 = 1,624 naloxone trainee-months during which 18 administrations of naloxone occurred: and so around 1.1 naloxone administrations per 100 naloxone-trainee-months.

I’d like to know the answers to questions 1) to 9) so that the approximate calculations, as shown in italics, could be made less approximate by actually using the evidence-base from NTA’s publicly-funded pilot studies.


Conflict of Interest: SMB is co-principal investigator for the prison-based pilot randomized controlled N-ALIVE Trial which aims to quantify the effectiveness of Naloxone-on-release for reducing drugs-related deaths soon after release for eligible adult prisoners with a history of heroin injection.