Elective Surgery Before, During and After the COVID-19 Pandemic in England 2015 - 2022: A Database Study

Study design

The aim of this study is to examine the effect of COVID on access to planned NHS-funded surgery overall and at NHS vs independent hospitals and in Priority groups 3 and 4. We do this by plotting the monthly number of procedures before, during and after the COVID lockdowns and recording the percentages in different groups over these three time windows. We report the study in accordance with STROBE criteria.(4)

Setting

The data used for this study were extracted from the Hospital Episodes Statistics (HES) database that contains all emergency and elective patient admissions, outpatient appointments and A&E attendances funded by the NHS in England. All procedures funded by the NHS were included, irrespective of whether they took place in NHS or independent hospitals.

Participants

Admissions for elective surgery (spells) were included if they met the following criteria:

• One of the operation codes from Priority group 3 or 4 listed in Supplementary Table S2 in primary position.

• Operation date between 1^st April 2015 and 30^th April 2022.

• Not a repeat of a particular operation for a patient.

• Patient age on admission between 18 and 120.

• Admission method is elective (not emergency).

Patients with unknown sex or age, and patients with inconsistent data (e.g. date of death preceding operation date) were excluded.

Variables

Information extracted about each surgical procedure is itemised in Table 1.

Data sources

Data were extracted from the HES database. Our study was based on adult patients (aged ≥18 years old) with selected operation codes. We extracted data over three time windows: April 2015 – March 2000 (pre-COVID); April 2000 – March 2021 (COVID); and April 2021 –April 2022 (post-COVID).

We identified the procedure codes for the most frequently performed procedures from each of these two categories. The list of procedures used in this study, along with their OPCS4 codes, are given in Supplementary Table S2.

Bias

In this study, we focus on the most frequently performed procedures from Priority groups 3 and 4. These procedures may not be representative of all procedures in these groups, but nevertheless provide a picture of the most common operations. We explore the issue of bias further in the discussion.

Study size

In this study, we have included all procedures that satisfy the inclusion criteria. For Priority 3, there were 3,378,611 procedures and for Priority 4, there were 607,259 procedures.

Statistical methods

For groups of patients treated at NHS sites and ISHP sites, we obtained:

1) Number of surgeries across the above three time-windows

2) Types of surgeries across the three time-windows.

Since the time-windows are of unequal duration, we made all comparisons per unit of time and include the monthly mean number of operations. We plotted the results across the time windows to allow visual inspection and describe the overall trends in operations.

We describe differences between pre and post-COVID proportions of surgeries by ethnic and socioeconomic group. However, it would not be appropriate to test the statistical significance of the difference between these proportions as we do not have data on the underlying proportions in the population.

In addition, we describe, across time-windows the ages of patients undergoing elective surgery, the comorbidity groups (using Charlson comorbidity scores), 90-day death rates and 90-day emergency readmission rates.

Data access and cleaning methods

Investigators at University Hospitals Birmingham (UHB) have access to pseudonymised, patient-level HES data for the purposes of this study. Data were cleaned using the inclusion criteria listed above and aggregated for reporting purposes. This observational study was registered with the local Clinical Audit Department (Clinical Audit Registration and Management System number 17421). Data were used in line with the data sharing agreement with NHS Digital.

Participants

The disbursement of participants is shown in Figure 1.

• Download figure
• Open in new tab

Figure 1: Disbursement of participants

We identified 2,624,440 Priority 3 surgeries in time-window 1; 238,990 in time-window 2; and 515,181 in time-window 3 (Table 2). For Priority 4, the corresponding figures are 488,502; 41,886 and 76,871 (Table 3). Counts of operations performed over time can be visualised in Figure 2.

• Download figure
• Open in new tab

• Download figure
• Open in new tab

Figure 2: Operation numbers in NHS and ISHP hospitals over time and by priority ranking

Time-trends

Figure 2 shows the changing number of Priority 3 and 4, NHS-funded operations taking place in ISHPs and NHS hospitals over the three time windows. The data track the massive drop off in operations at the high points of the pandemic, followed by incomplete recovery.

Proportions Treated in NHS vs Independent Hospitals Over Time-Windows

In both priority groups, we can see that there is a shift between operations in NHS hospitals vs independent sector hospitals in the pre- and post-COVID time windows. In Priority 3, 79% of operations were carried out by NHS providers before COVID and this falls to 63% post-COVID. Similarly, the percentage falls from 89% to 72% for Priority 4 operations.

Figure 3 shows the same operation numbers indexed to April 2015. These plots show that independent hospitals more than recovered after COVID and NHS hospitals are yet to return to pre-COVID levels. For Priority 3 operations, there were twice as many in independent hospitals at the start of 2022 as there had been in April 2015 and over three times as many for Priority 4. In NHS hospitals, however, there were half as many for Priority 3 and 4 as there had been in April 2015. This is in the direction hypothesised in the introduction (Question 1).

• Download figure
• Open in new tab

Figure 3: Indexed Operation numbers in NHS and ISHP hospitals over time and by priority ranking

Key results

We found that, overall, operation rates have not recovered to their pre-COVID level. This has obvious implications for the reduction of waiting times and health effects for the population. Our data are silent on the reasons for failure for surgery rates to recover. Conceivably, this is simply part of a secular trend and some support for this idea comes from the trend line observed in Figure 2. Alternatively, and we think more plausibly, the COVID pandemic has done lasting structural damage that include staff shortages observed elsewhere in the economy.(5, 6)

It is clear that the independent sector has absorbed some of the increased demand for elective surgical services. Elsewhere, we have shown that outcomes for NHS-funded patients are better following operations in the independent sector compared to NHS hospitals after propensity score matching and across different risk levels.(7)

We find clear evidence that under conditions of high demand, higher value procedures are awarded increasing priority. Whether this trend has gone far enough is a topic worthy of further study, but it is reasonable to suppose that as the number of these procedures increases, so the cases selected are those which are causing the patient greater distress. Insofar as this is true, the utility gain per operation will rise as the numbers of such operations performed declines – the formal argument is laid out elsewhere.(8)

With regard to socioeconomic and ethnic groups, comparisons must be interpreted with caution because we do not know how denominators may be shifting in the population. In the particular issue of ethnicity, the proportions change because of rapidly changing willingness to declare an ethnic group (Tables 2 and 3). However, there is one signal that stands out from noise and that cannot plausibly be attributed to shifts in underlying proportions in the population – the spike in YAG eye operations. These are done mainly in White people and in independent hospitals. This enigmatic finding warrants further explanation.

Limitations

There are always limitations when working with routinely collected healthcare data, mainly that they are not collected for the purpose of audit, service evaluation or research. There are possible coding errors and missing data, however our analysis is purely descriptive and shows the trends over a very large number of operations.

Interpretation

The English NHS is responding to the crisis of rising waiting times by intervening at the supply side and the demand side. At the demand side, the government is producing an app that can direct patients to providers with the lowest waiting lists.(9) A possible problem with this policy is that it assumes that supply can be rapidly expanded to meet increased demand in organisations with currently low waiting times. At the supply side, the government is promoting High Volume Low Complexity Hubs, whose resources are ring-fenced to provide separate packages for elective operations in elective facilities.(10) A call has been announced by the National Institute for Health and Care Research (NIHR) to conduct a thorough investigation of these Hubs. In due course we will have evidence on whether Hubs can reduce waiting times overall. In the meantime, it would appear that the independent sector is absorbing some of the load of elective surgery, but with less discrimination between more and less urgent cases.

Generalisability

The data used for this study are for the most frequently performed operations in each priority group. Therefore, our conclusions are strictly generalisable across these operation types.