Suspected cauda equina syndrome: no reduction in investigation, referral and treatment during the COVID-19 pandemic

Abstract

Introduction

Compression of the cauda equina can lead to bladder, bowel and sexual dysfunction with lower limb pain, numbness and weakness. Urgent surgical decompression aims to prevent progressive neurological deficit. Symptoms of cauda equina syndrome (CES), such as back pain, sciatica and bladder dysfunction are common in the population, but the majority of those investigated do not have radiological cauda equina compression. However, a missed diagnosis can have significant medical, social and legal consequences. We investigated the effect of the COVID-19 pandemic on presentation and management of suspected CES.

Methods

This retrospective cohort study analysed referral, investigation and treatment of CES in a regional neurosurgical centre during the initial COVID-19 surge between March and May 2020 compared with March to May 2019.

Results

Referrals for suspected CES were similar during the COVID-19 pandemic (n = 275) compared with 2019 (n = 261, p = 0.596) despite a significant (19%) decrease in total emergency neurosurgical referrals (1248 in 2020 vs 1544 in 2019, p < 0.001). Nineteen (7%) of the suspected CES referrals underwent decompression in 2020, similar to 16 (6%) in 2019 (p = 0.867). There were no differences in outcomes or complications and no evidence of delays in presentation or treatment.

Conclusions

Unlike other emergency neurosurgical conditions, the number of referrals for suspected CES and the percentage of referrals with radiological cauda equina compression were unchanged during the COVID-19 pandemic. The persistence of CES referrals when many people stayed away from hospital highlights the distress and worry caused by suspected CES and its symptoms to both patients and healthcare providers.

Introduction

Cauda equina syndrome (CES) encompasses a range of clinical presentations including bladder, bowel and sexual dysfunction, loss of saddle sensation, back and leg pain, and lower limb neurological symptoms.¹ Incomplete resolution of symptoms can lead to significant disability,² and diagnostic and treatment delays may result in worse outcomes and high legal costs.^2,3 Although the incidence of CES from acute lumbar disc prolapse is less than 1 in 100,000,⁴ the population prevalence of lower back pain can be up to 60%,⁵ and that of urinary incontinence up to 50% in women,⁶ so CES is much more commonly considered as a putative diagnosis.⁴ Fewer than 20% of MRI scans for suspected CES show cauda equina compression,^4,7 but the burden of investigation must be weighed against the consequences of a missed or delayed diagnosis.

The World Health Organization declared COVID-19 a global pandemic on 11 March 2020.⁸ Measures to contain the pandemic led to changes in provision and access to healthcare, work, and lifestyle. Decreases in emergency department (ED) attendances,⁹ and in presentations and treatment for non-COVID-19 emergency conditions, were reported.^10,11 This may reflect reduced incidence of disease or an altered threshold for presentation due to fear of COVID-19 infection or altered social situations.^9–11 We investigated the effect of the COVID-19 pandemic on presentation, investigation, treatment and outcomes of CES with the aim of better understanding the significance of and threshold for suspected CES presentations and referrals among patients and healthcare professionals.

Methods

We analysed emergency neurosurgical referrals to the Department of Clinical Neurosciences, Edinburgh, the regional referral centre for CES in south-east Scotland. The period from 1 March to 31 May 2020 was compared with the same period in 2019. This period was chosen to examine the effects of COVID-19 and associated lockdown measures on healthcare provision and help-seeking behaviour. The first COVID-19 case was confirmed in Scotland on 1 March 2020, cafés, pubs and restaurants closed on 20 March 2020, the UK lockdown began on 23 March 2020, and initial lockdown measures were eased in Scotland on 29 May 2020.

We analysed total numbers and proportions of emergency neurosurgical referrals, CES referrals, radiologically confirmed CES cases and other emergency neurosurgical subspecialty referrals. We compared referral sources, treatment, outcomes and complications in patients undergoing surgery. Data sources were consecutive daily emergency neurosurgery referral documentation and patient electronic health records. Complete data were available for both periods. This audit was approved by the department’s clinical governance team.

Count data were reported as median with interquartile range and compared using Mann–Whitney U tests. Proportions were compared using chi-square tests or Fisher’s exact tests where counts were below ten. All analyses were performed in R version 4.0.0.

Results

Between 1 March and 31 May 2020, 275 referrals (median: 3/day) were received for suspected CES, compared with 261 (median: 2/day, p = 0.6) between 1 March and 31 May 2019. By contrast, total emergency neurosurgical referrals were significantly lower in 2020 compared with 2019 (1,248 (median: 12/day) vs 1,544 (median: 18/day), p<0.001), for all neurosurgical subspecialties except spine (Table 1). Non-CES spine referrals were lower in 2020 (166, median: 1.5/day) compared with 2019 (232, median two/day, p = 0.004). Total admissions to neurosurgery were lower in 2020, but the number of admissions for suspected CES was similar (Table 1).

Table 1 .

Comparison of cauda equina syndrome and neurosurgical referrals and admissions before and during the COVID-19 pandemic

	2019	2020	p-value
Suspected CES referrals
Total	261	275	0.596
Per day	2 (1–4)	3 (1–5)
CES/neurosurgery referrals	261/1544 (17%)	275/1248 (22%)	0.006
CES admissions
Total	67	62	0.419
Per day	1 (0–1)	0 (0–1)
CES/neurosurgery admissions	67/371 (18%)	62/233 (27%)	0.058
Radiologically confirmed CES cases
Total	16	19
Per day	0 (0–0)	0 (0–0)	0.948
CES cases/CES referrals	16/261 (6%)	19/275 (7%)	0.867
Neurosurgery referrals
Total	1544	1248	<0.001
Per day	18 (13–20)	12 (8–17)
Neurosurgery admissions
Total	371	233	<0.001
Per day	4 (2–6)	2 (1–3)
Referral subspecialty
Vascular
Total	219 (14%)	165 (13%)	0.007
Per day	2 (1–3)	2 (1–2)
Trauma
Total	468 (30%)	368 (30%)	0.001
Per day	5 (4–7)	4 (3–5)
Spine
Total	493 (32%)	441 (35%)	0.176
Per day	5 (3–8)	4 (2–7)
Hydrocephalus
Total	131 (9%)	88 (7%)	0.006
Per day	1 (1–2)	1 (0–1)
Tumour
Total	212 (14%)	171 (14%)	0.032
Per day	2 (1–3)	2 (1–3)
Other
Total	21 (1%)	15 (1%)	0.334
Per day	0 (0–0)	0 (0–0)

Data are for March to May 2019 and March to May 2020. Both total numbers and the median count with interquartile range per day are given. CES = cauda equina syndrome

A similar proportion of referrals resulted in radiologically confirmed CES during both periods (19/275 (7%) in 2020 vs 16/261 (6%) in 2019, p = 0.87). Forty (17%) patients were referred twice in 2020 (either pre- and post-magnetic resonance imaging (MRI), or by primary and secondary care) and 34 (15%) were referred twice in 2019. Of patients referred as suspected CES, 19 of 235 (8%) had radiological cauda equina compression in 2020 and 16 of 227 (7%) in 2019 (p = 0.806). Figure 1 illustrates key COVID-19 time points alongside the decrease in neurosurgical referrals, but sustained CES referrals.

Figure 1 .

Time series of neurosurgical referrals, cauda equina syndrome (CES) referrals and CES cases during the COVID-19 pandemic compared with the previous year. Data are seven-day rolling averages of counts plotted for each 24-hour period. NS = neurosurgery.

Figure 2 shows that a lower proportion of referrals were made by general practitioners during the pandemic (19% in 2020 vs 28% in 2019, p = 0.023). This was associated with a non-significant increase in referrals from EDs (57% in 2020 vs 49% in 2019, p = 0.074) and physiotherapists (10% in 2020 vs 7% in 2019, p = 0.173). All patients with clinical and radiological CES underwent surgical decompression during both periods. Most patients underwent surgery within 48 hours of decision to operate (89% in 2020 vs 94% in 2019, p = 1.0; Figure 3). There was no significant difference in the length of time from symptom onset to presentation (p = 0.433) and a non-statistically significant small increase in median length of stay (4 (range 2–11) days in 2020 vs 3 (range 2–11) days in 2019, p = 0.061; Figure 3). Outcomes at discharge were not significantly different (resolution of symptoms: 6/20 (30%) in 2020 vs 9/16 (56%) in 2019, p = 0.176; required catheter: 3/20 (15%) in 2020 vs 2/16 (13%) in 2019, p = 1.000; required walking aid: 5/20 (25%) in 2020 vs 3/16 (19%) in 2019, p = 0.709). Intraoperative durotomy occurred in two of twenty (10%) cases in 2020 and one of sixteen (6%) cases in 2019. There were no reoperations, complications, readmissions to neurosurgery or mortality within 30 days in either period. There were no postoperative medical complications or known COVID-19 infections within 30 days. Three patients had COVID-19 swabs taken, all of which were negative. One was taken postoperatively after discharge because of symptoms and was negative. Two were taken preoperatively. One had a negative result prior to the operation commencing, and the other negative result was only available postoperatively.

Figure 2 .

Source of cauda equina syndrome (CES) referrals during the COVID-19 pandemic compared with the previous year. ED = emergency department, GP = general practitioner, PT = physiotherapist.

Figure 3 .

Timings of presentation, decompression and discharge during the COVID-19 pandemic compared with the previous year. Graphs include 16 confirmed cases of cauda equina syndrome (CES) in 2019 and 19 confirmed cases of CES in 2020.

Discussion

The numbers of referrals, admissions and operations for suspected and radiologically confirmed CES were similar during the peak of the COVID-19 pandemic compared with the same period in 2019. This contrasts with a decrease in other neurosurgical referrals. A similar number of patients underwent surgery during both periods. There was no evidence of delayed presentation, delayed time to theatre or increased morbidity.

These findings contrast with reports of decreased ED attendances⁹ and decreased cases of myocardial infarction¹⁰ and stroke¹¹ during the COVID-19 pandemic. Attendances for associated symptoms such as chest pain also fell.¹² The decrease in healthcare episodes for non-COVID-19 conditions during the pandemic may reflect lower activity levels due to lockdown restrictions, a higher threshold for seeking care, lack of healthcare provision or worry about exposure to COVID-19 in the healthcare setting.^9,10,12 Acute lumbar disc prolapses are often atraumatic and precipitated by bending and twisting so the incidence of CES would not be expected to be affected by lockdown. However, similarly, tumour presentations would not be expected to change with lockdown, yet these did decrease (171 in 2020 vs 212 in 2019, p = 0.032; Table 1). Patient and healthcare provider concerns regarding significant consequences of the symptoms experienced, or a missed or delayed diagnosis^2,3 may contribute to consistent referral numbers, but referrals for other neurosurgical conditions with equally significant symptoms and consequences did decrease.

Most people investigated for CES do not have radiological cauda equina compression. In our department, the emergency imaging protocol includes a T2 sagittal MRI of the cervicothoracic spine if no structural cause for the symptoms is identified on lumbosacral imaging. This means the 3–9% of patients with an alternative structural cause such as demyelination or spinal fracture are usually diagnosed during admission.^13,14 However, 70–80% of patients have severe, acute onset pain, bladder, bowel or sexual dysfunction but have normal or non-explanatory imaging.^13,14

Three recent studies have investigated these patients in more detail.^13–15 Two studies divided patients into those with nerve root compression and those without.^4,14 In a prospective study, there was a step-wise increase in levels of chronic pain, pre-existing bladder dysfunction, particularly stress incontinence, panic attack at symptom onset and positive signs of a functional neurological sensory or motor disorder from those with cauda equina compression compared with those with nerve root compression and those without compression.¹³ All studies found higher rates of functional disorder comorbidities in those presenting with suspected CES than would be expected even in a population with chronic back pain.^13–15 Mechanisms of CES-like symptom production are likely to be related to pain, panic, exacerbation of prior bladder dysfunction, medications such as opiates causing direct neural inhibition of the bladder, locally via urethral sphincter contraction, and centrally by inhibitory signals affecting higher order bladder control, and possible triggering of acute functional neurological weakness or sensory loss.¹³

During the initial COVID-19 surge, home working with poor ergonomic arrangements, home improvement activities, poor sleep and increased psychological distress may all have contributed to increase pain.¹⁶ Increases in functional disorders have been reported following warfare and terror attacks in both civilians and those involved more directly, but appear less common after natural disasters.¹⁷ Situations in which there is lack of control, unpredictability and poor social or financial support may trigger or exacerbate pain and functional disorders.^16,17 At a time when the public have reported being frightened and discouraged from attending hospital, it emphasises the high levels of distress and genuine nature of those presenting with CES-like symptoms, whether or not they have a purely structural explanation.

The proportion of patients with radiological cauda equina compression was similar in 2020 and 2019 (8% vs 7%), but lower than estimates from a recent meta-analysis (19%)⁴ and UK-wide audit (15.6%).⁷ In Edinburgh, 30 of 326 (16%) referred patients underwent decompression for CES during the six-month UK-wide audit in 2016–2017, in keeping with the UK average.⁷ Updated UK guidelines for investigation of CES in December 2018 broadened the indications for emergency MRI scanning to include those with bilateral sciatica without visceral or sphincter symptoms, and those with back pain and/or sciatica rather than back pain with sciatica.¹⁸ We purposely compared the COVID-19 period with March to May 2019 as this was after introduction of the revised guidelines. Lower rates of radiological cauda equina compression than previously published may reflect a lower threshold for investigation since the updated guidelines.

We did not find any evidence for a significant delay in presentation or treatment of CES, although the small numbers with CES preclude adequately powered assessments of small differences. The pandemic did not appear to increase operative morbidity or affect outcomes. The small increase in length of stay from a median of 3 days to a median of 4 days was not statistically significant and may not be clinically relevant. None of our patients developed symptomatic COVID-19 in the 30-day follow-up period, although only one patient with symptoms was tested, and asymptomatic patients were not tested in Scotland during the study period. COVID-19 testing for asymptomatic surgical patients began at the end of May 2020. Testing undertaken in patients’ local health boards was not necessarily available for review.

This was a single centre study and radiologically confirmed CES occurs infrequently, so there were few CES cases. However, we analysed over 1,000 neurosurgical referrals and 200 suspected CES referrals in each period. We only assessed referrals made to the on-call service and did not include outpatient referrals. Our analysis assumes that all presentations worrying the healthcare provider or patient for suspected CES were referred to us during both periods. As referrals did not decrease during the COVID-19 pandemic it is unlikely that our findings are due to decreased onward referrals with sustained presentations to other healthcare providers, but we cannot exclude the possibility that presentations increased and the threshold for onward referral decreased. This single centre study has the advantage of full ascertainment of all referrals and data directly comparable with the previous period with no difference in referral pathways or MRI provision. Patients are referred to our centre with suspicion of CES rather than with a confirmed diagnosis of CES (in contradiction to UK guidelines)¹⁸ due to the lack of local MRI provision. Our centre has the only access to 24/7 MRI in south-east Scotland, with some regional hospitals providing a weekday in-hours service with limited service at weekends, and some local hospitals not always able to provide even daytime urgent access to MRI.

Conclusion

Referrals for suspected CES were similar during the COVID-19 pandemic compared with the previous year, contrasting with the decrease seen in other neurosurgical emergency referrals. This underscores the significance of the clinical syndrome of symptoms for those both with and without radiological cauda equina compression.

Competing interests

JS reports honoraria from UpToDate. JS and PFXS carry out expert witness work in personal injury and negligence claims. JW, PB, and IH have nothing to disclose.

Funding

No funding was received for this study. JS is supported by a National Research Scotland Clinical Fellowship.

Author contributions

JW, PB and PFXS conceived and designed the study. JW acquired the data and drafted the manuscript. JW, PB, PFXS, JS and IH analysed and interpreted the data, revised the manuscript for important intellectual content, and approved the final version. JW and IH are accountable for the accuracy and integrity of the work.