- New data shows 1 February reduction in agency pay caps drove weekly breaches above 45,000 – an increase of 10,000 a week
- Breaches began to decline again at the start of March
- NHS Improvement estimates agency caps saved £290m between October and February
The number of NHS agency shifts per week that breached new pay caps rose by more than 10,000 in the first three months after they were introduced, new data shows.
According to figures released by NHS Improvement, the health service breached the caps 35,662 times in the week beginning 23 November 2015. Weekly breaches peaked three months later – on the week starting 22 February 2016 – at 46,491.
However, breaches began to decrease at the end of February, falling to 42,843 by the week commencing 7 March, when the data series finishes.
The capped rates, which have been introduced to curb spiralling NHS agency bills, include “break glass clauses” that allow trusts to breach the cap in order to protect patient safety. Since the caps were introduced they have been progressively reduced, first on 1 February and again on 1 April.
NHS Improvement, which is responsible for implementing the caps, argues that the override figures have to be seen in the context of the cost reductions the agency cap is delivering. For the first time, the regulator has put a figure on this for the months since the policy’s introduction, estimating that the controls saved £290m between October and February.
An NHSI spokeswoman told HSJ agency spend over this period was £1.5bn, against expected spending of £1.8bn based on previous trends.
It also argues that the figures show breaches increase with each new set of controls introduced, but then steadily decrease as the market adjusts.
After the initial caps were introduced in November, the number of weekly breaches fell sharply, reaching a low of 21,227 in the week beginning 28 December.
However, the number of breaches then began to rise again – before the caps were reduced – and by the end of January had almost returned to November’s level.
When the caps were reduced on 1 February the number of breaches shot up to 45,686. It remained at that level or higher throughout the month, then reduced slightly to 42,843 by the week of 7 March.
The NHSI spokeswoman said it expected the number of breaches will have increased again when the caps were lowered on 1 April, but “once the new caps embed themselves the override figures will steadily decrease once again”.
Asked how the increase in breaches throughout January fitted with this hypothesis, she said the “January figures we believe were a bounce back after a very low Christmas period”.
The data NHSI has published to date only provides a partial picture of the effects the caps are having. The absence of weekly data on rota fill rates makes it hard to assess whether the caps are driving an increase in rota gaps, with a corresponding risk to patient safety.
When NHSI’s predecessor Monitor published agency cap data in January, the Nuffield Trust called for it to also publish detailed ward level data on fill rates.
Nuffield Trust senior policy analyst Sally Gainsbury said: “If Monitor want to convince people that the caps are working we need to know that not only are hospitals spending less money on agency staff but also still maintaining quality and patient safety.”
Agency spend figures for the first 11 months of 2015-16, also released by NHSI, show expenditure peaked between July and September at £951m. It then fell to £902m for October to December, the quarter when the caps were first introduced.
For January and February, expenditure was £285m and £287m respectively. These figures were lower than in any other month of 2015-16, at a point in the year when winter pressures would normally be expected to drive increased spending. This suggests the introduction of the cap did succeed in curbing agency spending growth in-year.
The overall figures for weekly cap overrides have been verified. However, the figures for different categories of agency staff – also shown in the chart above – have not been verified and may be subject to change.