Prioritizing delivery of cancer treatment during a COVID19 lockdown - the experience of a clinical oncology service in India


 INTRODUCTION: A COVID19 lockdown in India posed significant challenges to the continuation of radiotherapy (RT) and systemic therapy services. While several COVID19 service guidelines have been promulgated, implementation data is yet unavailable. We performed a comprehensive audit of the implementation of services in a clinical oncology department.METHODS AND MATERIALS: A departmental protocol of priority-based treatment guidance was developed and a departmental staff rotation policy was implemented. Data was collected for the period of lockdown on outpatient visits, starting, and delivery of RT and systemic therapy. Adherence to protocol was audited, and factors affecting change from pre-COVID standards analyzed by multivariate logistic regression.RESULTS: Outpatient consults dropped by 58%. Planned RT starts were implemented in 90%, 100%, 92%, 90% and 75% of priority level 1-5 patients. While 17% had a deferred start, the median time to start of adjuvant RT and overall treatment times were maintained. Concurrent chemotherapy was administered in 89% of those eligible. Systemic therapy was administered to 84.5% of planned patients with 33% and 57% of curative and palliative patients receiving modified or deferred cycles. The patient’s inability to come was the commonest reason for RT or ST deviation. Factors independently associated with a change from pre-COVID practice was priority level allocation for RT and age and palliative intent for systemic therapy.CONCLUSIONS: Despite significant access limitations a planned priority-based system of delivery of treatment could be implemented.


Introduction
In the wake of the COVID19 pandemic, India went into a strict lockdown on March 24, 2020. This included the abrupt cessation of all types of public and private transport except for de ned essential services 1 .
The sudden lockdown resulted in oncology services in India facing a crisis of decision-making and delivery of care.
From March 2020, guidelines on cancer treatment, and risk-strati ed care had started emerging2. These generally suggested modi cation or deferment of treatment if considered safe. Although oncology services around the country started adopting one or more of the recommendations 3-5, the available literature is limited to consensus guidelines and surveys primarily based on western healthcare infrastructure. Implementation of a planned approach from a system that does not have structured state funding for travel and treatment has not yet been audited or reported.
We put in place a detailed protocol to prioritize care pathways using available evidence, biological rationale and published consensus statements (Appendix I). We present here an audit of our services from March 24 through May 16, 2020. The focus of this audit was the implementation of treatment delivery amongst our patients.

Methods
Departmental triaging and treatment protocols The departmental policy (Appendix I) was based on the treatment priorities in uenced by treatment intent and disease biology. Prioritization for radiotherapy (RT) was primarily based on the recommendations of the NHS UK and divided cancer cases into ve levels2. For patients on systemic therapy, we used a modi ed priority system based on the curative or palliative intent of therapy. We maintained a prospective database of all cases whose treatment was deferred during this period.

Data sources
We obtained patient visit data between January 1st -May 16th for the years 2019 and 2020 from the electronic hospital information system (HIS). The HIS and oncology information system (ARIA, Varian Medical System, Palo Alto, USA) was queried to obtain information on patient characteristics and treatment delivery patterns between March 24th -May 16th, 2020. Study data were collected and managed using REDCap electronic data capture tools6, 7.
Statistical analysis R8 and Python 3 were used for statistical analysis. We used the Chi-square test and the Kruskal Wallis test was used for statistical testing of differences in frequencies and continuous variables respectively.
Multivariate modeling was used to identify the factors predicting deviation of radiotherapy and chemotherapy from pre-COVID protocols (Appendix II). Multivariate analysis was done using logistic regression, where the presence of any deviation was considered as the independent variable. Model predictors were added linearly and no interactions were assumed. Odds ratios, 95% con dence intervals, and p values are presented. A p-value of <0.05 is considered statistically signi cant.

Outpatient visits
Within the period of Jan 1-May 16, 5291, and 5090 patients had outpatient visits in 2019 and 2020 respectively. There were 12325 outpatient consults between Jan 1 to May 16, 2020, compared to 13140 in the same period in 2019. While there were 1983 (25%) excess outpatient visits in the rst 12 weeks of 2020 as compared to 2019, there was a sharp drop in patient visits induced by the lockdown in the 13th week. The average weekly follow up visits in the four most common site-groups of breast, lung, head and neck, and prostate cancers dropped by 65%, 49%, 50%, and 76% respectively.

Radiation therapy
During the period of lockdown, there were 305 patients who were planned for starting RT. Of these, 262 were able to start on treatment between the period of 24th March to 31st May 2020 (Table 1). Breast (27%), head and neck (23%), and lung (17%) were the commonest sites(Appendix II, Table 1). Of the 145 patients in priority levels 1, 2, and 3, 132 (91.0%) could begin their treatment during the lockdown.
In 125 (47.7%) patients, either adjuvant RT after surgery or de nitive RT after induction chemotherapy was delivered. Of these, 54 patients had priority level 1, while 69 had priority level 5. The median time to start RT was 40 days (range: 14 -69 days) after surgery or the last cycle of chemotherapy. A delay of 6 weeks in starting RT was observed in 17 (31.5%) patients. In all but one of these patients, the delay beyond 6 weeks was due to restrictions in patient travel, nances, or due to delayed attendance in our hospital after a surgery done in another hospital. Among priority 5 patients undergoing adjuvant radiotherapy, the median gap between RT and the last cycle of chemotherapy or surgery was 53 days (range: 16 -110 days).
The median RT plan turnaround time (TAT) was 7 days. Among priority level 1 patients, plan TAT exceeding 14 days was observed in 3 (2.6%) patients. Plan TAT did not exceed 14 days in any of the priority 1 patients on adjuvant or post-induction RT. All patients in priority level 2 started on the same day of planning.
Concurrent chemotherapy was indicated in 65 (32.0%) of the 203 patients. Amongst these 58 (89.2%) patients were started on chemotherapy. Chemotherapy could not be delivered per protocol in 26 patients (27.1%) due to treatment-related toxicities. Of these, 7 patients had chemotherapy stopped early, 1 had a dose reduction while 9 had missed one or more cycles of chemotherapy.
Delivery of RT 429 patients underwent RT during lockdown (262 new and 167 ongoing) ( Table 2, Appendix II). By deferring starts of priority level 5 patients, on-treatment numbers reduced to an average of 129 per day during the period of lockdown from 172 earlier (Fig 1, Appendix II). A total of 8 patients had breaks, and six patients could not complete their planned treatment.
Of the six, two were unable to come for further therapy, while the remaining progressed or died during treatment (unrelated to COVID). Overall treatment time was prolonged by more than 3 days in 14 patients (5 priority level 1).

Downtime
We faced a great challenge with downtime, with one of the 4 treatment units down for technical reasons on 24 days of the 59 working days (inclusive of Saturdays). However timely shifting of patients to alternative units was done which is re ected in the overall treatment time.
Brachytherapy 25 patients with gynecological cancers were planned for brachytherapy. Of these, brachytherapy could be delivered in 17 patients (11: cervical cancer, 5: endometrial cancer, 1: vault recurrence). A scheduled brachytherapy source exchange had to be deferred during the lockdown as a result of which 8 patients (4 cervical cancer and endometrial cancers each) were referred outside for brachytherapy after May 1, 2020.
In the 17 patients who received brachytherapy, treatment was completed in 15 patients. 2 patients with endometrial cancers were unable to come for the last fraction of vaginal brachytherapy. The total duration of treatment for patients who underwent treatment during this period was less than 56 days for all except 1 patient.
Factors affecting deviation from pre-COVID usual radiotherapy practice Factors affecting deviation from practice are shown in Appendix II tables 3 and 4. Figure 2 (panel A) shows that the only factor which was independently associated with deviation from pre-COVID protocol was the priority level. Compared to priority level 1, priority level 5 had an odds ratio of 4  Table 2 shows the implementation of chemotherapy in these patients. About one-third of curative intent chemotherapy and close to 60% of patients on palliative chemotherapy had some form of deferral from planned dates. The median duration of delay was longer in palliative patients (28 days vs 7.5 days, p = 0.002). Deferrals in curative patients were equally related to disease or toxicity related causes, inability to attend due to lockdown and physician recommendations. Deferrals in patients on palliative systemic therapy were more commonly due to physician recommendation (56.4%). In a smaller proportion of patients, there was a change of chemotherapy schedule, mainly related to reduced intensity. In only 26 (12%) patients chemotherapy was stopped completely, and 18 of these patients were on palliative treatment. Toxicity-related stoppage or deferral was uncommon.
Factors affecting deviation from pre-COVID chemotherapy practice Factors affecting deviation from usual practice are shown in Appendix II tables 5 and 6. Figure 2 (panel B) shows that the two factors independently associated with deviation from pre-COVID chemotherapy practice were increasing age (OR between 3rd and 1st quartiles 3.48, 95% CI 1.71 to 7.07, p = <0.01) and palliative intent chemotherapy (OR 3.03, 95% CI 1.28 to 7.14, p = 0.01), which re ects our modi ed intent during the lockdown.

Discussion
When the lockdown was imposed nationwide in India on March 24, the state of West Bengal had 9 con rmed cases, and one death due to COVID19, increasing to 2532 cases and 232 deaths on May 16, 20209. While this reduced casualties from COVID19, lockdown had wide-ranging effects on other healthcare services.
The magnitude of the effect of the COVID 19 induced lockdown on cancer care is emerging globally 10, 11. Data from Prime Minister Jeevandayi Aroyga Yojna (PM-JAY), an universal health insurance scheme, show claims related to oncological care fell by nearly 64% during the lockdown 12. Systemic chemotherapy deferrals and delayed start of new patients on chemotherapy were responsible for this decline. Similar reports of disruption in oncological care delivery and its impact have emerged from other healthcare delivery systems in Germany13, Japan 14, Italy15, 16 and the UK17, 18.
To ensure service continuity, services have adopted a system of staff rotations during this crisis 4, 16, 19, 20. Treatment prioritization enabled us to continue delivering safe treatment with reduced staff. Our team was able to reach out to patients with scheduled appointments and provide guidance based on priority levels. This is re ected in a greater drop in follow-up patient visits in breast and prostate cancer patients.
In terms of RT services, we took the decision of not postponing or interrupting RT in patients who were already undergoing treatment. In hindsight this decision was proven correct as to date there is no sign that the epidemic is abating in India despite the lockdown. However, by deferring new starts for priority level 5 patients, we were able to reduce the new starts. This ensured that manpower for planning and treatment could be strategically re-deployed by rotation based 50% attendance of radiation therapists and medical physicists/dosimetrists to prevent delays and ensure safety in delivering full services for priority level 1-3 and symptomatic level 4 patients.
For priority level 1-3 patients, this strategy succeeded in implementing more than 90% of planned starts. Deferments and incomplete treatments were linked primarily to patients being unable to come for treatment. Concurrent chemotherapy was also successfully implemented in the majority. The only factor predicting a change from pre-COVID RT practice was the priority level assigned, which matched with our intent.
The department continued to offer specialized procedures which are highlighted by the fact that one patient successfully underwent Total Body Radiation as a part of the conditioning regimen for bone marrow transplant 21. Additionally, complex planning techniques were used as indicated and no change in planning technique was done. For example, all patients of breast cancer continued to be treated with cardiac sparing using Deep Inspiration Breath Hold 22. This is unlike the experience in some western centers, where similar complex procedures were suspended 23. Signi cant changes in our dose fractionation schedules were not required as our pre-existing departmental policy was to use hypofractionated radiotherapy wherever it was safe24-26.
The two factors which predicted a deviation from the usual pre-COVID chemotherapy delivery were age and use of palliative chemotherapy -both of which were in line with our proposed departmental protocol (Appendix I). For curative intent patients, on the other hand, a change in dose density was offered. This is re ected in the duration of the deferral of chemotherapy. We also offered G-CSF based prophylaxis to all patients.
Similar experiences from other centers are yet to be reported. No episodes of transmission of COVID 19 from staff to patient or vice versa were observed. Daily pretreatment screening and appropriate counseling of patients and staff may have contributed. Furthermore, with the support of the administration, we were able to ensure that staff could travel from far-ung areas in the city. We must acknowledge the fact that the patients understood the importance of their disease and it's treatment and were motivated to continue on the treatment during this period 27.
The downstream effects of lockdown, in terms of delayed diagnoses or delayed access for newly diagnosed patients in the community, will only become apparent with further follow-up. Modeling results suggest a signi cant increased risk of death due to delayed treatment and diagnosis in this population28. Declarations Acknowledgments: We would like to acknowledge all the members of the radiotherapy department especially, our departmental coordinators Mr Sougata Banerjee and Ms Nandita Das who ensured that treatment delays and deferrals were minimized during this crisis. We would also like to thank all patients who choose to repose their trust in us and continue with the treatment despite facing signi cant hardships. No extramural or intramural funding was received for this study.

Tables
A statement of ethics approval: The study received a waiver from the Institutional Review Board / Ethics committee at our institution as this is an audit (Waiver no: EC/WV/TMC/33/20) A statement of participant consent: In view of this being an audit without any patient contact or interventions there was a consent waiver from the Ethics Committee.  Figure 1 Showing the change in case load in the OPD In the period corresponding to the rst, second, and third phases of the lockdown (weeks 13-15, 16-18, and 19-20) for major cancer site groups.