Colorectal Cancer Surgery at a Regional Cold Site Centre During the SARS-CoV-2 Pandemic: A Comparison With Prepandemic Practices
ABSTRACT
Introduction
Stockholm, Sweden had among the highest excess mortality rates during the first wave of the SARS-CoV-2 pandemic. In March 2020, an entire hospital implemented an infection-free cold site infrastructure and scaled up imperative mainly colorectal cancer surgery, for the Stockholm health care region. The study evaluated the efficacy of the concept.
Methods
Strict preadmission screening and prophylactic interventions were implemented. The outcomes of the cold site cohort (30th March–21st June 2020) were compared to a prepandemic control. Population-based data on colorectal cancer patients were collected from the Swedish Colorectal Cancer Registry.
Results
There were 107 patients in the cold site cohort and 109 in the control cohort. None developed COVID-19 within 30 days postoperatively. The proportions of pelvic and laparoscopic surgery were similar. The cold site cohort had more potentially avoidable stomas (32/94 (34%) vs 19/91 (21%), p = 0.0442). In the region, the proportion of colorectal cancer resections at the cold site increased from prepandemic 12.9% (58 of 448) to 43.3% (94 of 217) (p < 0.001).
Conclusion
A SARS-CoV-2-cold site can be kept infection-free by screening, allowing increased volumes of colorectal cancer surgery with outcomes comparable to prepandemic practice. The cold site model should be more widely adopted in future similar scenarios.
Summary
This study reports on the implementation of an entire cold site hospital for imperative colorectal surgery, in particular colorectal cancer, during the SARS-CoV-2 pandemic. By comparing outcomes to prepandemic practices, we demonstrated that the cold site hospital was effective in preventing infection and safely increasing surgical volume many-fold.
1 Introduction
Like many health systems around the world, the regional health system in Stockholm, Sweden, was overwhelmed by the coronavirus disease 2019 (COVID-19) pandemic in spring 2020. Patients suffering from moderate to severe COVID-19 filled existing hospital wards and critical care units beyond capacity. Additional facilities had to be repurposed and healthcare workers reassigned from other services.
As a consequence, acute care hospitals had a severely diminished capacity for elective care, often even for emergency care other than for COVID-19. In addition, patients who did undergo elective major surgery during the first wave were found to have a mortality rate above 20% if they developed COVID-19 in the perioperative period, itself a not insignificant risk [1].
The concept of cold site hospitals, therefore, emerged early in the pandemic, describing centres where special measures were taken to entirely exclude infected patients and staff from the site, allowing delivery of imperative health care, including major surgery [2, 3]. In some countries, cold sites were widely implemented. In England and Wales, 54% of the hospitals had access to a cold site for colorectal cancer surgery, which was deemed pivotal in maintaining the capacity for treatment [4]. Limited published data on urological and various surgical procedures indicate that the cold site model is safe [5, 6].
The Stockholm Region health authority tasked Ersta Hospital with increasing its number of imperative surgery to meet the demands of the wider region from March 30, 2020, and a SARS-CoV-2 cold site structure was implemented to deliver this safely. In this study, we investigated the safety and efficacy of performing major colorectal surgery, mainly for colorectal cancer, at a SARS-CoV-2 cold site during the first wave in this region, which was severely affected by the pandemic. Furthermore, we compared perioperative outcomes in this setting with those achieved prior to the COVID-19 pandemic. Finally, pandemic and prepandemic volumes of colorectal cancer surgery are compared between cold and hot sites at a population-based level.
2 Materials and Methods
This observational, retrospective study compared two similarly sized single centre cohorts of patients who underwent colorectal resections for benign and malignant disease before and during the first wave of the COVID-19 pandemic. Hospital volumes were compared in a population-based cohort including all patients who underwent bowel resection for colorectal cancer during the same time periods in the Stockholm region, with 2.4 million inhabitants. This study was approved by the Swedish Ethical Review Authority (2021-02291), and reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
2.1 Setting
Ersta Hospital is a 65-bed university hospital providing high-volume elective care for gastrointestinal diseases, including colorectal cancer and inflammatory bowel disease. It has five operating theatres and three intensive care beds. Although it does not have an emergency department, it provides emergency admissions and surgery for its own patients at all hours. In the Stockholm region, colorectal cancer surgery is performed at five hospitals. During the pandemic, all hospitals except for Ersta Hospital served as hot site hospitals caring for SARS-CoV-2 infected patients in wards and intensive care units.
2.2 Participants
Two cohorts of patients who underwent imperative elective or expedited small bowel, colon, or rectal surgery at Ersta Hospital were identified from an international prospective clinical registry of perioperative outcomes, the Enhanced Recovery After Surgery Interactive System (EIAS) [7]. Patients who underwent stoma procedures (diversion or reversal through peristomal incisions) were not included in the EIAS or in the present study. Patients aged < 18 years were excluded.
The “Cold site” cohort consisted of patients who underwent surgery from 30th March to 21st June 2020. This time interval of 12 weeks was the period of maximum demand on the healthcare system in Sweden during the first wave of the pandemic, as measured by overall hospital admissions and intensive care admissions [8].
The “Control” cohort consisted of all patients fulfilling the same criteria from 1st January to 30th June 2019. A longer time period was chosen to approximately match the sizes of the two cohorts. Outcome data were collected during the first 30 postoperative days.
In addition, data were analysed for all patients in the Stockholm healthcare region who underwent bowel resection for colorectal cancer during the two time periods.
2.3 Intervention
The Stockholm Region advised all hot site hospitals to refer SARS-CoV-2-negative patients requiring imperative colorectal surgery to Ersta Hospital when demand exceeded their capacity.
The entire hospital implemented a range of measures by 30 March 2020 to achieve two objectives: to prevent infection in the hospital and to increase case volumes many-fold. (Table 1). These measures were partly adopted from the first clinical guidance issued regarding colorectal surgery during the pandemic.
| Objective | Intervention | Methods |
|---|---|---|
| Infection prevention | Preadmission SARS-CoV-2 screening | Telephone screening |
| CT chest screening | ||
| Nasopharyngeal PCR screening | ||
| Surgical technique | Additional personal protection equipment on airway management | |
| Low-pressure laparoscopy | ||
| Decreased power in energy instruments | ||
| Gas capture | ||
| Limit number of staff in theatre | ||
| Safe manyfold case volume increase | Halt nonimperative surgery | |
| Transfer additional specialist surgeons to cold site | ||
| Truncate preoperative pathway | Multidisciplinary team conference in referring hospital | |
| Preoperative consultations during admission one day prior to surgery | ||
| Truncate postoperative pathway | Increased usage of end and diverting stomas | |
| Increased use of rehabilitation facilities for recovery |
To achieve the first objective, to minimise the risk of infection of patients and staff in the hospital, all patients were screened prior to admission. Patients were initially screened by telephone, those with upper respiratory tract symptoms or an increased temperature were advised to follow national recommendations regarding suspected infection, and their operation was deferred 3 weeks or more. Patients without such symptoms attended a separate screening clinic in the hospital the day before admission for diagnostic testing. Given that there was a shortage of nasopharyngeal polymerase chain reaction tests (PCR) during the first weeks of the pandemic, screening was initially performed via computed tomography of the chest (CT) [9]; later, both CT and PCR were performed. The oxygen saturation, core temperature, circulating white cell count, and C-reactive protein level were also measured in all patients prior to admission, and any findings out of range were assessed by a clinician. Patients were not admitted to the hospital until screening was complete and negative. Key measures to minimise the transmission of SARS-CoV-2 in any patient with false-negative screening results included additional personal protection equipment during airway management, low-pressure laparoscopy (10 mmHg) and gas capture for the elimination of pneumoperitoneum (using standard suction and filters). These measures were partially based on clinical recommendations emerging at this time [10, 11]. Meals were served bedside rather than in the ward dining hall, and liberal PCR testing for SARS-CoV-2 was used in cases of postoperative pyrexia or respiratory symptoms. Screening of asymptomatic staff beyond national guidelines was not performed.
To achieve the second objective, to increase case volumes many-fold without affecting the quality of care, all nonimperative surgeries were stopped, and the surgical workforce was reinforced by consultant surgeons from other hospitals in the region. Patients referred for resection of colorectal adenocarcinoma were reviewed at a cancer multidisciplinary team (MDT) conference in the referring hospital, according to national clinical guidance. Patients were scheduled for surgery directly according to the clinical decision at the initial MDT conference, but all radiological images were reviewed by a radiologist and surgeon before surgery. Similarly, patients referred for imperative surgery for inflammatory bowel disease (IBD) were listed for surgery without formal review at the IBD MDT conference.
Instead of attending surgical and anaesthetic outpatient clinics in the weeks prior to surgery, all patients were admitted to the hospital in the afternoon prior to surgery for assessment. Clinical pathways, including neoadjuvant and adjuvant oncological therapies and preoperative optimisation for IBD were unchanged, as was the institutional enhanced-recovery after-surgery pathway. The use of end and diverting stomas was increased to decrease postoperative morbidity in accordance with international clinical guidance [12]. A closer collaboration with regional rehabilitation centres was arranged, aiming for further reduction in hospital stay.
2.4 Outcomes
The degree to which the first objective, hospital infection prevention, was achieved was assessed by the proportion of patients diagnosed with SARS-CoV-2 during their hospital admission and during the first 30 days after surgery. All patients were followed up by phone 30 days postoperatively. Any symptoms of SARS-CoV-2 infection or PCR test results after discharge from the hospital were obtained. The second objective, a safe manyfold increase in case volume, was assessed by comparing the postoperative morbidity and postoperative length of hospital stay between the two cohorts as well as the case volumes in the corresponding time periods. Morbidity was measured as the Clavien-Dindo grade of the most serious complication in each patient in the first 30 postoperative days [13], with no complication coded as 0. Furthermore, intestinal stoma rates and proportions of operations completed laparoscopically were also compared.
Finally, to assess regional changes in referrals and shifts in surgical volume to the cold site hospital, we compared the proportions of patients who underwent resection for colorectal adenocarcinoma at the cold site and in the entire healthcare region during the two time periods.
2.5 Data Sources
Data were obtained from the EIAS database with incomplete or missing data validated against medical records. Information on preoperative COVID-19 screening was obtained from medical records. The coded study data were stored in a secure electronic database at Ersta Hospital. From the Swedish Colorectal Cancer Registry, population-based data on diagnosis, date of surgery, and operating centre were collected for colorectal cancer patients.
2.6 Statistical Methods
All analyses were performed via JMP 16 for MacOS (SAS Institute Inc, Cary, NC, USA). Continuous variables are reported as medians (ranges). Categorical variables are reported as frequencies and percentages. Any missing data are indicated and were not replaced. The Wilcoxon rank-sum test and Fisher's exact test were used to compare groups.
3 Results
3.1 Cohorts
During the period from 30 March to 21 June 2020, 107 patients underwent small bowel, colon, or rectal surgery (the cold site group). This corresponds to an increase of 133% compared with that 1 year earlier, when 46 patients underwent such operations from 30 March to 21 June 2019. The control group was expanded to include all 109 patients who underwent small bowel, colon or rectal surgery from 1st January until 30th June 2019. There was no loss to follow-up, and all patients were analysed.
The groups had similar sex, age, and comorbidity distributions (Table 2). In the cold site group, the dominant indication for surgery was colorectal adenocarcinoma, whereas in the control population, half the cohort underwent surgery for benign conditions such as diverticulosis, inflammatory bowel disease, and other benign diseases (Table 2). Sixty-eight patients in the cold site group (64%) were referred from hot-site hospitals.
| Cold site group (n = 107) | Control group (n = 109) | |
|---|---|---|
| Sex n (%) | ||
| Female | 52 (49%) | 55 (50%) |
| Male | 55 (51%) | 54 (50%) |
| Age (years) | 68 (18–89) | 66 (28–93) |
| American Society of Anesthesiologists Classification | ||
| I | 3 (3%) | 9 (8%) |
| II | 82 (77%) | 85 (78%) |
| III | 22 (21%) | 15 (14%) |
| Indication for surgery | ||
| Colorectal adenocarcinoma | 94 (88%) | 58 (53%) |
| Inflammatory bowel disease | 11 (10%) | 13 (12%) |
| Diverticular disease | 2 (2%) | 14 (13%) |
| Other benign disease | 0 (0%) | 22 (20%) |
| Other malignant disease | 0 (0%) | 2 (2%) |
3.2 Screening Outcomes
All 107 patients in the cold site group underwent screening for SARS-CoV-2; all by chest CT and 80 by chest CT and nasopharyngeal PCR. The CT scan was negative in all patients. One of the 80 PCR tests was positive. This patient had no symptoms suggestive of SARS-CoV-2 infection. The operation was rescheduled 3 weeks later, at which time repeat PCR was negative.
No patient was diagnosed with a SARS-CoV-2 infection during admission or within 30 days after surgery.
3.3 Surgery
The type of surgical procedure differed between the groups (p < 0.001), with more right-sided colon resections in the cold site group, although the proportion of abdominal versus pelvic resections did not differ (p = 0.287, Table 3). The choice of surgical approach did not differ (p = 0.687, Table 3), nor did the rate of conversion from laparoscopic to open surgery (4 (5%) vs. 2 (2%), p = 0.394). Stoma rates did not differ between groups (p = 0.097); however, when operations inherently associated with an end stoma (abdominoperineal excision of the rectum and proctocolectomy) were excluded, the rate of diversion was found to be increased in the cold site group (32/94 (34%)) compared with the control group (19/91 (21%), p = 0.044). The surgical management of rectal cancer did not differ between the groups, with eight patients undergoing low anterior resection and nine undergoing abdominoperineal resection in the cold site group compared to 10 and 10 in the control group (p = 0.858).
| Cold site group (n = 107) | Control group (n = 109) | p | |
|---|---|---|---|
| Surgical procedure | 0.0002 | ||
| Right-sided colon resection | 43 (40%) | 28 (26%) | |
| Left-sided colon resection | 35 (33%) | 33 (30%) | |
| Total colectomy | 8 (7%) | 4 (4%) | |
| Reversal of Hartmann's operation | 0 (0%) | 8 (7%) | |
| Low anterior resection | 8 (7%) | 10 (9%) | |
| Abdominoperineal excision of rectum | 10 (9%) | 14 (13%) | |
| Proctocolectomy | 3 (3%) | 4 (4%) | |
| Other | 0 (0%) | 8 (7%) | |
| Surgical details | |||
| Abdominopelvic resection | 21 (20%) | 28 (27%) | 0.287 |
| Operation completed laparoscopically | 83 (78%) | 87 (80%) | 0.687 |
| New stoma formed | 45 (42%) | 37 (34%) | 0.0970 |
| 30-day morbidity and mortality (Clavien-Dindo) | 0.249 | ||
| 0 | 68 (64%) | 69 (63%) | |
| I | 12 (11%) | 6 (6%) | |
| II | 19 (18%) | 25 (23%) | |
| IIIa | 3 (3%) | 1 (1%) | |
| IIIb | 4 (4%) | 8 (7%) | |
| IVa | 0 (0%) | 0 (0%) | |
| IVb | 0 (0%) | 0 (0%) | |
| V | 1 (1%) | 0 (0%) | |
In 22 of the 107 operations (21%) in the cold site group, a surgeon from another hospital operated or assisted. This did not occur in the control group.
3.4 Postoperative Recovery
The postoperative length of stay did not differ between the cold site group and the control group at 5 (2–22) and 5 (2–48) nights, respectively (p = 0.119). A similar proportion of patients were discharged to a rehabilitation facility, 33 of 107 (31%) versus 23 of 107 (22%, missing data n = 2) (p = 0.119).
The distribution of complications (Table 3) and the rate of severe complications (Clavien-Dindo grade 3b or higher) (5% vs. 7%, p = 0.408) did not differ between the cold site and control groups. There was one death within 30 days in the cold site group. This occurred on postoperative day 20, 17 days after discharge home, and postmortem examination revealed acute myocardial infarction.
3.5 Changes in Referrals
All four hot site hospitals had a decreased capacity for surgery during the intervention period, and colorectal cancer patients were referred to Ersta Hospital for surgery. In the Stockholm region, the proportion of colorectal cancer patients operated on at Ersta Hospital increased from prepandemic 12.9% (58 of 448) to 43.3% (94 of 217) (p < 0.001) during the pandemic (1st January until 30th June 2019 vs. 30th March to 21st June 2020).
4 Discussion
In the present study, we describe the experience of a unique cold-site hospital in Sweden, a country with one of the highest excess mortality rates during the first wave of the pandemic [14].
We demonstrate that with thorough screening of all patients prior to admission, as well as staff observing strict infection control measures, the hospital effectively blocked in-hospital COVID-19, with no cases diagnosed during the extended postoperative period. Preadmission patient screening consisted of screening questions, clinical examination, laboratory tests, CT imaging of the chest, and nasopharyngeal PCR.
Furthermore, the measures taken to increase case volumes were successful, with an increase in the number of major colorectal surgeries by 133% during the first wave of the pandemic compared with the same time period the year before. Among all colorectal cancer patients in the Stockholm region, the proportion operated on at the cold site increased from 12.9% to 43.3% when comparing the prepandemic and pandemic time periods. Postoperative outcomes were not affected by this marked increase in operating volume; the postoperative length of hospital stay and the severity of postoperative complications did not differ between the pandemic and control time periods. Some differences in surgical practices between the pandemic and control time periods were noted. As expected, indications for surgery during the pandemic were exclusively imperative. They included mainly cancer but also several cases of inflammatory bowel disease with uncontrolled disease or dysplasia. The case mix in the control cohort reflects everyday colorectal surgical practice, with approximately half of the caseload being benign disease, including functional conditions. The operations performed also differed; for example, the reversal of Hartmann's operation was not performed during the pandemic. These observations provide an indication of the diagnoses and types of operations that were delayed during the pandemic. They also highlight that imperative colorectal surgery does not consist only of cancer surgery; many patients with inflammatory bowel disease, in particular, have rapidly progressive disease that may be a threat to life in the medium term should their surgical care be delayed, an issue that has frequently been overlooked during the pandemic by nonspecialists [15].
The rate of minimally invasive surgery remained at approximately 80% during the pandemic, and the conversion rate did not increase. The safety of laparoscopy compared with laparotomy was questioned in the early phase of the pandemic due to safety concerns of theatre staff inhaling aerosolised tissue fluid containing the virus from infected patients [10, 12]. An assessment was made that the advantages of minimally invasive surgery in terms of improved postoperative recovery outweighed these theoretical considerations, particularly as additional measures were taken to reduce carbon dioxide release straight into the operating theatre. Subsequent publications have reported different assessments across the world in relation to laparoscopy; in England, the rate of laparoscopic resections for colorectal cancer fell from 59% before the pandemic to 25% during the first wave [16], whereas in Denmark, no change in the use of minimally invasive surgery was observed [17].
The rates of potentially avoidable stoma increased during the pandemic (34% vs. 21%). Raising the threshold for a primary anastomosis was a clinical choice on the basis of international clinical guidance to minimise the risk of severe complications requiring intensive care and prolonged hospital stays [12]. Similar findings were reported in studies of international and English practice [16, 18]. An increased rate of abdominoperineal resection to low anterior resection for rectal cancer was not found presently.
Following the first wave of the pandemic, some data on the cold site provision model were published. Kasivisvanathan et al described outcomes in the first 500 patients treated during the first wave at a COVID-19 cold site in London [3]. Screening was performed by telephone call and selective CT or PCR testing. Ten patients (2%) were diagnosed with postoperative COVID-19 and none died. Glasbey et al reported outcomes in an international cohort of patients undergoing surgery for a range of malignancies in hospitals described as running COVID-19-free surgical pathways (separating of theatre and facilities from general hospital areas) compared with patients in hospitals that did not implement such pathways [2]. Postoperative respiratory complications, including postoperative SARS-CoV-2 infections, were approximately halved within COVID-19-free surgical pathways.
This study has several limitations. The number of patients who had symptoms compatible with SARS-CoV-2 infection on telephone screening and were therefore rescheduled without further screening was significant but not recorded. Outcome data for patients who underwent colorectal resections at hot spot hospitals in the region were not available. This study concerns perioperative practice, and oncological therapies and long-term outcomes for patients who underwent surgery for malignant disease were not analysed. A recent nationwide analysis suggests that oncological outcomes of colon cancer treatment were not affected overall by the pandemic in the region [19].
Sources of bias were minimised by including all consecutive patients aged 18 years or older during the two time periods. Seasonal variation is a source of bias, but such variation occurs mainly during the winter and summer holiday periods, mainly occurring outside both time periods. The present results are generalizable to comparable health systems affected by viral epidemics of similar severity and epidemiology.
5 Conclusion
In conclusion, the present results demonstrate that rapid implementation of an entire cold-site hospital was possible and effective during the first wave of the COVID-19 pandemic, both in terms of ringfencing the site from infection and in terms of significantly scaling up volumes of surgical care delivered without affecting perioperative outcomes. These findings support wider implementation of cold-site hospitals during future similar events on the basis of strict preadmission testing and truncated clinical pathways to scale up imperative surgery including colorectal cancer resections.
Acknowledgements
The authors would like to thank their colleagues in the Department of Surgery and Anaesthesia at Ersta Hospital for scaling up and maintaining safety in the delivery of major surgery in challenging circumstances, as well as the Stockholm Region health authority for their support.
Open Research
Data Availability Statement
Available on request from the corresponding author.
References
Synopsis
This study reports on the implementation of an entire cold site hospital for imperative colorectal surgery, in particular colorectal cancer, during the SARS-CoV-2 pandemic. By comparing outcomes to prepandemic practices, we demonstrated that the cold site hospital was effective in preventing infection and safely increasing surgical volume many-fold.
