Levels of harm

Dear Sir

I am very keen on Patient Safety and feel it is a very important topic which needs addressing seriously and sensibly.
I was therefore distressed to see de Wet and Bowie's paper quoted in Table 2 as estimated levels of harm in the community. In fact this paper looked at both acute and primary care. Most harm and most serious harm in the paper occurred in secondary care. The paper also confuses error and harm. Please look at my comments on-line at the Postgraduate Medical Journal. In fact the primary care harm rate was less than 1% when properly analysed.
Please withdraw your paper and rewrite to take these facts into account.

Best wishes

Chris Johnstone

PS you are not the only people to have made this mistake. The paper is poorly written and presented. I am attempting to have it withdrawn to prevent errors such as this. I am sorry for the time and expense you will have to go to to rewrite your paper, maybe you could contact the PMJ and ask them to withdraw this paper to prevent other well-meaning people making the same error.

Dear Dr Johnstone,
Thank you so much for drawing this to our attention. I will look into the reference to de Wet and Bowie's paper with the author of the scan and come back to you within the week.

It's very important to us that we don't misrepresent any of the research contained within the scans. I appreciate the information.

Kind regards,
Jonathan

Dear Dr Johnstone,

After discussing this over with colleagues and the author of the scan, we believe that de Wet and Bowie article should remain within the scan, but that text making reference to the scan will be clarified to clearly reflect the nature of the evidence within the paper. Our reasons for this are outlined below.

The Health Foundation research scans aim to draw together the empirical research published on a topic. Unlike our Best Evidence reviews or a systematic review, they do not weigh or judge the methodological merits of each piece of research. The paper from de Wet and Bowie is clearly relevant to the topic of levels of harm in healthcare.

The eligibility criteria for this scan state that any reviews or articles containing primary research or secondary review were considered for inclusion. Clinical audit definitely falls into this category. Empirical research categorised as research, evaluation or audit within a research ethics committee would all be eligible.

The current reference to the harm rate in de Wet and Bowie’s research is taken directly from the from de Wet and Bowie’s own conclusions. Reflecting upon your very helpful feedback, we have decided to rephrase references to findings from this research within the scan in order to emphasise the distinction between harm and error, and between harm/error captured in primary care records rather than error/harm taking place in primary care. This will be updated shortly.

I would be happy to talk this through in more detail if you would like.

Kind regards,
Jonathan