A post-covid elective recovery scenario

A post-covid elective recovery scenario

21/04/2021
by Rob Findlay

The latest planning guidance for England’s NHS envisages a service restricted by covid until the end of September. So activity will probably be lower than usual.

At the same time, many of the patients who stayed away from the NHS during covid are expected to start coming back. So demand will probably be higher than usual.

Higher demand and lower activity means a growing waiting list, and growing waiting times. How bad might it get?

And after covid, we all want to make waiting lists safe for patients with cancer and other urgent conditions, and then get back to 18 week waits again, what might the recovery plan look like?

Using the published RTT data as the starting point, I thought I would try to answer those questions.

Longer – safer – shorter

This Tableau Public viz allows you to select a trust and specialty, and see what this year and the subsequent recovery might look at, under a particular scenario (other scenarios are, of course, equally straighforward to model). The analysis was done in Gooroo Planner and modelled in three phases.

From April to September 2021, activity progressively steps up from 70 to 85 per cent of 2019-20 levels as described in the planning guidance. At the same time, the pent-up demand is assumed to come in: that’s all the missing referrals since March 2020, deducting 15 per cent removals, and assuming 80 per cent of the rest will return. This is a worst-case scenario, and this is why Referral-Decision demand is so high right at the start.
From October 2021 to March 2025, the priority is to make the waiting list safe by getting patients to diagnosis quickly. The objective is to reduce Referral-Decision waiting times to 8 weeks, without letting Decision-Treatment waits blow out beyond 40 weeks. So Referral-Decision activity steps up significantly, which in practice will probably be helped along by new outpatient pathways.
From April 2025 to December 2028, the priority is to reduce Decision-Treatment waiting times while keeping Referrral-Decision waits short. At the end of 2028, the aim is to achieve 18 weeks from referral to treatment with room to spare.

The analysis takes into account the shortfall between activity and demand that existed in most of the NHS before covid, as well as the knock-on effects of outpatient activity on the admitted waiting list. When estimating future demand I have used 2019 as the baseline, which is probably a reasonable guess at the moment, but later this year (if you are doing your own planning with Gooroo Planner) it will probably be better to switch to using proper forecasting of demand and performance.

The analysis does not include the following (although it could):

Trend growth in demand has not been factored in, so as not to obscure the effects of covid. (When Gooroo Planner is used by trusts and commissioners, they usually include demographic and other trend demand growth assumptions).
Non-elective pressures have not been modelled here, and in real life the pressure on non-elective beds in particular would be taken into account. (Local Gooroo Planner models usually include non-elective services, and bed occupancy can be based on achieving acceptable levels of risk using Gooroo Beds.)
Non-consultant-led activity is excluded because it is excluded from the source RTT data (local Gooroo Planner models usually include this), as are pressures on workforce, backlog maintenance, social care, primary care, and all the other pressures that will need to be addressed alongside elective waiting times.
Some local services are missing from the data, because a complete set of data could not be extracted from the published RTT tables. (Local Gooroo Planner models usually take stage-of-treatment data direct from the local data warehouse which avoids this issue.)
The published data is monthly, so that has been mapped to weekly data to take account of seasonality in Gooroo Planner. (Local Gooroo Planner models would automatically measure or forecast based on the patient-level data so that no such estimation is needed.)

Making comparisons

You may be wondering how to compare the activity levels with your existing capacity. I am afraid that the published RTT data does not make this easy, but here are some pointers. (None of these issues arise in a local implementation of Gooroo Planner because stage-of-treatment data comes from the usual data warehouse or PAS tables.)

The September 2021 activity levels are 85 per cent of 2019-20 activity, so you can use that to guide the eye.
Decision-Treatment activity should match consultant-led elective inpatient and daycase admissions.
Referral-Decision activity will not bear a close resemblance to outpatient or diagnostic activity, unfortunately, because the published RTT data does not capture all of those stages; more detail here.

Finally, you may be wondering about the discontinuities in the Decision-Treatment waiting time charts. Gooroo Planner calculates the best waiting time that can be sustained safely, given the level of clinical priority, the size of the waiting list, and the rate that patients are being added to it (among other things); you can find the detailed research behind this calculation here. When the addition rate shifts around, Gooroo Planner usually smooths those shifts to reflect the way that waiting lists absorb variation, but it can only do that within a model. When you are linking three different models together, as we are here, the discontinuities between one model and another come out into the open. I can only suggest that you allow your eye to smooth them out over a period of about 4 months, which is what would happen in real life as bulges and dips work through the waiting list.

Your own scenarios

If you would like to make your own covid backlog recovery plans based on your own data and scenarios, either for a trust or across a system, then please get in touch to discuss implementing Gooroo Planner.

Return to Post Index