The Sunday Telegraph’s page one lead story on 13 January said: “Seventeen NHS hospitals have dangerously low staffing levels, according to rulings by the official safety watchdog [ie: the Care Quality Commission]”.

The story was given added veracity by quotes from the secretary of state − “no excuse for hospitals not to provide adequate staff” − and similar ones from his Labour shadow and the chief executive of the Patients Association. The story was picked up widely, for example running as a major story on BBC Online.

‘We have to be very careful about how we describe and communicate unrepresentative data about single wards’

The only problem is that the story was untrue… as anyone could have found out had they contacted the hospitals concerned or looked at the level of patient impact in the detailed inspection reports on the CQC website, as the Foundation Trust Network has now done. When you do this, a very different picture emerges to that portrayed in The Sunday Telegraph.

It’s important to say up front that four of the 17 hospitals we contacted acknowledged that the CQC had reasonable concerns about staffing levels at the point when the inspection took place, though the hospitals would argue that “dangerously understaffed” is not an accurate reflection of those concerns.

In addition, these issues had already been addressed by the time the article was published and in some of these cases the CQC had already reinspected and confirmed that they no longer had concerns. So it would have been helpful had the article pointed out that this information was historical. The story’s use of “17 NHS hospitals have [note the present tense] dangerously low staffing low levels, “latest inspections” and “inspections took place as recently as November” convey the opposite impression.

But that’s not the main problem with the story. Because in the other 13 of the 17 it was factually incorrect to state that recent CQC data showed that the hospital had “dangerously low staffing levels”. It’s impossible at short length to give chapter and verse on each of the 13 but there are three main, common sets of reasons as to why the description of “dangerously low staffing levels” is a gross exaggeration.

1. No distinction between minor, moderate and major patient impact

The first is that the data provided by the CQC, via a Labour information request, only set out whether the hospitals were compliant or non compliant on the CQC “staffing” inspection outcome. The information completely failed to distinguish between whether the patient impact of non-compliance was minor, moderate or major (risk to patient impact is what a CQC inspection assesses and these are the three grades of impact).

None of the CQC reports on the trusts recorded a risk of major patient impact. They were all minor or moderate impacts or there was no summary judgement because they pre-dated the CQC adoption of this classification. What does this mean in reality? Two examples will suffice:

  • In one hospital, the ward had a single member of staff who was off sick at very short notice. The ward was short-staffed and the ward sister had been unable to find a replacement but was happy to provide cover herself. After two hours’ conversation between the CQC inspector and the ward sister in her office, she insisted that she needed to return to patients. Unsurprisingly, when the inspector then talked to patients, they observed that the ward had been understaffed that day.
  • In another, the hospital was in the process of recruiting a replacement activity therapist. Their job was to prevent patients from becoming bored and to help them with interesting activities. When the CQC inspector interviewed patients, one patient who had been on the ward before noted the (temporary) absence of an activity therapist.

In both cases, these translated into “minor patient impacts” (as anyone reading the relevant CQC inspection reports would have seen) because no patient safety was at risk. In both these cases, and the others where a minor or moderate impact was noted, it is therefore patently incorrect to translate these impacts into “dangerously low staffing levels”. Only a CQC judgement of major concern could possibly justify that headline and the rest of the story.

2. Extrapolation of data

Another reason for the over-exaggeration is the extrapolation of data from one inspection report of one ward into a judgement on the hospital as a whole.

To be fair, buried in the middle of paragraph seven of the story is the short statement that “not every ward was visited”. But readers were encouraged to draw the conclusion that this was irrelevant as the chief executive of the Patients Association Katherine Murphy was quoted later in the story saying: “These are just the parts of the hospital that the inspectors have seen and there is no reason to think it will be any safer on the wards they have not visited”.

The truth, of course, is that there is no evidence to support this assertion and plenty to contradict it (for example, inspections of other wards in the hospitals concerned and subsequent inspections). Just one small ward recording a staffing issue with minor patient impact at one point was sufficient to turn the hospital into non-compliant as far as the CQC information was concerned, qualifying it for appearance in the article.

‘The FTN is not pretending poor quality of care doesn’t occur and that staffing levels are not an important factor in many of these cases’

Again, an example is instructive. In one hospital, a single sixbed escalation unit using a temporary overnight staffing pattern was inspected. There was a discussion between the hospital and the inspection team over the staffing level, with the inspection team agreeing that there was a legitimate professional debate over whether the staffing levels were right or not with the ultimate judgement a finely balanced one. This translated into a minor patient impact, again because no patient safety was at risk. But the Sunday Telegraph story clearly implies that all 291 beds in that hospital are on wards that are dangerously understaffed. This is clearly inaccurate.

The CQC is rightly in the process of changing its approach to inspecting acute hospitals as it recognises that inspections of single wards cannot provide accurate data about the whole hospital, or the wider trust to which it belongs.

But getting to a complete set of representative data will take time and, until we reach that point, we have to be very careful about how we describe and communicate unrepresentative data about single wards within acute hospital trusts.

3. Timeliness and accuracy of summary judgement data

A common theme among the trusts we spoke to, and a particular source of frustration, was that many had undergone subsequent inspections or reinspections of the same wards which gave their trusts a clean bill of health. This was not reflected in the summary “compliant/non-compliant judgement” data released by the CQC.

There was a sense that once a trust has been tagged “non-compliant” there was little it could to reverse this judgement and that CQC inspection data takes far too long to work its way through the CQC system − the latter is acknowledged by the regulator.

So, far from those hospitals being judged currently unsafe (or non-compliant) by the CQC, as the story implied, the majority had in fact been judged as safe by a later CQC inspection. The opposite of what the story said, if you apply the same evidence standard of using a single inspection to create a summary judgement. A simple phone call to the trust would have confirmed this but no one appeared to have bothered to ring any of the trusts involved.

Does all this matter? Nearly all the trust chief executives we contacted were angry and disappointed at the exaggeration and misrepresentation in the article. They had spent four days managing their boards, governors, commissioners and local media. Staff morale had understandably been dented with a real frustration about why their hospital was being treated as a political football.

Responsibility to staff and patients

Probably most concerning of all was the potential impact on patient safety. As one chief executive put it: “We have irrefutable evidence that when stories of this type occur, some patients who ought to be coming to our hospital don’t come because they are frightened. That’s where the real patient danger lies, not on our ‘dangerously understaffed wards’ which simply don’t exist”.

Just to be clear, the Foundation Trust Network is not pretending that poor quality of care doesn’t occur and that staffing levels are not an important factor in many of these cases. As outlined above, four of the 17 trusts acknowledged that the CQC inspection had reasonable concerns at the point when the inspection took place.

The FTN has also argued in an open letter recently that trust boards should be clearly held accountable, including through the media, for failures in care. But this has to be on the basis of the evidence, not exaggeration and supposition.

It’s going to be a difficult few months for the NHS once the Francis report has been published tomorrow. But everyone involved in commenting on the service − politicians, media, voluntary organisations − has a responsibility to staff and patients to respect the evidence. Exaggeration of the kind we saw on the front page of The Sunday Telegraph on 13 January should have no place in the debate at all.