Understanding the long-lasting impact of the COVID-19 pandemic on the diagnosis and management of lung cancer: a retrospective community hospital experience

Abstract

Background:

Cancer care was negatively affected during the COVID-19 pandemic. We sought to examine the long-term effects of the COVID-19 pandemic on patterns of lung cancer care among patients undergoing surgical resection at our thoracic surgery program.

Methods:

We conducted a retrospective cohort study of all patients who underwent lung resection for cancer at Michael Garron Hospital between 2018 and 2024. We categorized patients into 3 cohorts: prepandemic (2018 to 2019), COVID-19 pandemic (2020 to 2022), and postpandemic (2023 to 2024). We analyzed demographic characteristics, time to surgery, length of stay, multimodal treatment, and pathological stage.

Results:

We included 1090 patients. The time from consent to surgery was shorter in the pandemic cohort (26 v. 29 d; p < 0.01). Length of stay was also shorter during the pandemic and postpandemic periods (median 2 v. 3 d; p < 0.001). A higher proportion of patients presented with advanced-stage lung cancer (stage III) in the prepandemic cohort than in the other cohorts (15% v. 9%, p < 0.05).

Conclusion:

We found no significant delays in surgical care for lung cancer during the COVID-19 pandemic. Timelines to surgery may have been expedited owing to increased resource allocation. Through further study of adaptive mechanisms used during the pandemic, we can continue to provide increasingly effective and efficient lung cancer care.

Lung cancer is the leading cause of cancer death around the world, estimated to cause more than 20 000 deaths among Canadians in 2024 alone.¹ Although recent advances in immunotherapy show promise, surgical resection remains the primary curative treatment. Unfortunately, cancer care was negatively affected during the COVID-19 pandemic because health care systems were adapting to manage high volumes of patients with COVID-19, resulting in limited resources for other medical needs. The Canadian Institute for Health Information estimated a 30% decline in lung cancer surgeries during the early pandemic in 2020,² which corresponds with a predicted time of nearly 2 years to recover from the backlog of surgeries under optimal system capacity.³

Given this context, it is essential to characterize recent care patterns and identify systemic inefficiencies in surgical oncology. Zarnke and colleagues noted increased referrals for malignant disease but reduced patient attendance and prolonged consultation wait time through a Lung Diagnostic Assessment Program at a Canadian academic institution during the COVID-19 pandemic.⁴ Additionally, a large Canadian cohort study during the pandemic observed increased reliance on medical therapy or radiotherapy, even when surgery was the standard of care.⁵ These findings have been replicated on an international scale. Stuart and colleagues noted that patients diagnosed with non– small cell lung cancer in the United States in 2020 and 2021 tended to experience delays in treatment and presented at later clinical disease stage.⁶ A large-scale study of provincial administrative data estimated that each 4-week delay in cancer care treatment corresponds to a 6% increase in the risk of death. Concerningly, lung cancer was among the cancer types with the largest numbers of expected deaths in the aftermath of the COVID-19 pandemic in Ontario.⁷

In Canada, lung surgery is regionalized to thoracic surgical centres. Michael Garron Hospital (MGH) is a community hospital serving a catchment area of more than 400 000 patients. This study aimed to evaluate changes in patterns of lung cancer surgical care at MGH during and after the COVID-19 pandemic, focusing on treatment timelines and oncologic outcomes.

Methods

Patients

We conducted a retrospective cohort study using clinical data from patients who underwent lung resection for cancer at MGH between 2018 and 2024, identified via a surgical records database. Patients underwent clinical evaluation by thoracic surgeons for operability and resectability. Exclusion criteria included age younger than 18 years, receipt of neoadjuvant treatment, metastatic disease at the time of surgery, and a nonmalignant final diagnosis. We categorized patients into the following cohorts based on surgical dates: prepandemic (2018 to 2019), COVID-19 pandemic (2020 to 2022), and postpandemic (2023 to 2024).

Outcomes

The primary outcome was the pathological stage at surgery, defined by the AJCC Cancer Staging Manual, Eighth Edition.⁸ Secondary outcomes included timing between diagnosis, consent, receipt of multimodal therapy, surgery, and discharge date, as well as length of stay and surgical outcomes.

Data analysis

We extracted deidentified data through the institution’s electronic medical record. Demographic data included gender, age, and postal code (used to calculate distance from hospital). We performed statistical analysis using Stata software.⁹ We presented summary data as median and interquartile range or mean and variance. We performed the statistical analysis using χ² or Fisher exact tests for categorical variables. We used parametric and nonparametric methods for analysis depending on the distribution of the data (i.e., normal v. nonnormal distributions). We considered a p value less than 0.05 to be statistically significant.

Ethics approval

We obtained full research ethics board approval from the institution. Informed consent was waived given the retrospective nature of this study.

Results

We included 1090 patients (Table 1). The most common reason for exclusion of reviewed charts was that patients had metastatic disease and did not undergo surgery with curative intent. The median age of patients was 69 (interquartile range [IQR] 63 to 75) years. Based on the first 3 digits of patients’ postal code, the median distance travelled to the treating hospital was 22.3 (IQR 10.8 to 84.2) km. There were no significant differences in the number of patients who underwent biopsy among the 3 groups. Surgical technique was also similar among the cohorts: overall, 92% of patients underwent video-assisted thoracoscopic surgery (VATS), 5% had a thoracotomy, and 3% were converted intraoperatively from VATS to open surgery (Table 2).

Table 1.

Characteristics of patients who underwent lung resection for cancer (2018–2024), by period

Characteristic	No. (%) of patients;* period				p value
	Prepandemic	COVID-19 pandemic	Postpandemic	Total
All patients	378 (34.7)	383 (35.1)	319 (29.3)	1090	–
Age, yr, median	69	69	69	69	0.9
Sex, female	210 (55.6)	141 (36.8)	121 (37.9)	472 (43.3)	0.4
Time from consent to surgery, d, median	29	26	29	28	< 0.01
Distance from MGH, km, median (IQR)	21 (10–83)	23 (11–85)	24 (12–86)	22 (11–84)	0.03
Preoperative FEV1, median (IQR)	84 (70–98)	86 (73–99)	85 (72–97)	85 (72–98)	0.5
Preoperative DLCO, median (IQR)	80 (68–92)	81 (69–91)	82 (70–93)	81 (68–92)	0.5
Preoperative biopsy	343 (90.7)	181 (47.3)	111 (34.8)	635 (58.3)	0.09
PET SUV, mean ± SD	5.63 ± 4.92	6.78 ± 6.36	7.94 ± 6.10	6.70 ± 5.84	< 0.001

DLCO = diffusing capacity of the lungs for carbon monoxide; FEV₁ = forced expiratory volume in 1 second; IQR = interquartile range; MGH = Michael Garron Hospital; PET SUV = positron emission tomography standardized uptake value; SD = standard deviation.

^*Unless stated otherwise.

Table 2.

Preoperative considerations, surgical details, and pathological stage for included patients

Variable		No. (%) of patients;* period				p value
		Prepandemic	COVID-19 pandemic	Postpandemic	Total
PET SUV, median		3.9	4.7	6.55		< 0.001
Incision, %	VATS	92	93	91	92	0.7
	Open	5	4	6	5
	Conversion	3	3	3	3
Procedure	Lobe	82	83	85	83	0.7
	Segment	6	7	8	7
	Wedge	7	6	7	7
	Bilobectomy	1	1	2	1
	Pneumonectomy	2	3	2	2
Length of stay, d, median (IQR)		3 (2–4)	2 (2–3)	3 (2–4)	3 (2–4)	< 0.001
Final pathological stage	IA	213 (46)	149 (32)	100 (22)	462 (42)	< 0.001
	IB	96 (29)	105 (32)	130 (39)	331 (30)
	IIA	23 (24)	46 (48)	27 (28)	96 (9)
	IIB	17 (25)	25 (37)	26 (38)	68 (6)
	IIIA	60 (50)	31 (25)	31 (25)	122 (11)
	IIIB	3 (0.4)	3 (0.4)	2 (0.3)	8 (< 1)

IQR = interquartile range; PET SUV = positron emission tomography standardized uptake value; VATS = video-assisted thoracoscopic surgical.

^*Unless stated otherwise.

In unadjusted analyses using log-transformed distance to account for right-skewed data, we found a significant difference in travel distance across the 3 periods (p = 0.04). Compared with the prepandemic period, patients who underwent surgery in the postpandemic period travelled significantly farther, with a mean increase of 24% in travel distance, corresponding to approximately 27 km (95% confidence interval [CI] 0.04 to 0.39; p = 0.02). A similar trend was observed during the pandemic period, with an 18% increase in travel distance, equivalent to about 26 km, although this did not reach statistical significance (95% CI −0.01 to 0.33; p = 0.06).

The distribution of stage differed significantly across the 3 periods (χ² ₁₀ = 59.52; p < 0.001). Stage I disease (IA or IB) was most common across all periods, though a higher proportion of patients presented with stages II and III disease in the postpandemic group. A nonparametric test for trend also showed a significant increase in stage severity over time (z = 3.58; p < 0.001).

In a multivariable linear regression model, higher positron emission tomography (PET) standardized uptake value (SUV) of tumours and more advanced pathological stage were independently associated with shorter time from consent to surgery. For each unit increase in PET SUV, the time to surgery decreased by 0.34 days (β = −0.34, 95% CI −0.58 to −0.10; p = 0.005). Similarly, each 1-stage increase in pathological stage was associated with a 1.39-day decrease in time to surgery (β = −1.39, 95% CI −2.43 to −0.34; p = 0.009).

Discussion

This study demonstrates expedited surgical timeliness during the period of the COVID-19 pandemic, persisting shorter hospital stays postpandemic, and increased advanced-stage disease in recent years. The cohort in the pandemic period experienced shorter wait times between providing consent and their surgical date, as well as shorter length of stay in hospital. Several factors may have contributed to this at both the individual patient and systems level. Factors that may have expedited time to surgery include cancellation of elective nononcologic cases and adjusted operative booking patterns, which enabled prioritization of cancer cases and a relative increase in dedicated resources. In our study, the shorter length of stay in hospital persisted through the postpandemic period, but the time from consent to the surgical date increased to prepandemic levels. Fligor and colleagues¹⁰ describe the risks of delayed surgical care, noting significant increases in mortality among patients with early-stage lung cancer waiting more than 6 to 8 weeks for definitive surgery, whereas the National Comprehensive Care Network recommends surgery within 60 days of staging.^6,7 The median time from consent to surgery at our centre was approximately 4 weeks in both groups (26 v. 29 days; p < 0.01).

The results further demonstrate that in the current period (i.e., postpandemic) we are treating more advanced-stage cancers. Delays in presentation to medical attention for investigation of symptoms, less robust screening practices during and after the COVID-19 pandemic, and limited resources may have contributed to this trend. A cross-sectional study conducted during the early pandemic period identified uncertainty surrounding public recommendations for visiting public places and discomfort using transportation as primary factors contributing to unwillingness to attend medical appointments for cancer screening among the general population. ¹² Literature from more recent years is concordant with our findings: wait times between diagnosis and treatment were reduced during the pandemic years, but patients tended to have higher-stage disease at the time of diagnosis.^5,6,13 A retrospective study published in 2024 showed a significant increase in the T-stage during pandemic years compared with the pre- and postpandemic periods, with T2 to T4 disease comprising nearly 90% of cases between 2020 to 2022, compared with approximately 65% in the comparison cohorts.¹⁴ Canadian survey data on availability of diagnostic tests during the pandemic notes that most thoracic surgeons did not report a significant change in their ability to access computed tomography, but that bronchoscopy (87%), endobronchial ultrasonography (85%), and pulmonary function testing (63%) were less accessible during this time.¹⁵ Our study showed that both tumour PET SUV and higher pathological stage were independently associated with reduced wait time between consent and surgery, implying that patients with more advanced disease were being appropriately prioritized for surgery. These observations align with prior evidence that higher pretreatment PET SUV correlates with more aggressive disease and worse outcomes, including a meta-analysis in surgical non–small cell lung cancer showing higher SUV associated with increased risks of recurrence and death, and a recent surgical series showing that elevated maximum SUV is strongly associated with nodal upstaging in clinical stage IA disease.^16–19 These findings must be interpreted in the context of disruptions to health care delivery during a time of considerable social, economic, and political uncertainty from the pandemic.

Similar work has been reported by other Canadian groups. A single-centre study out of Montréal also found that there was a significant increase in later-stage presentations during pandemic years, and that the proportion of patients with M1 disease increased after 2020 compared with 2019 (> 50% compared with 31% prepandemic; p < 0.05).²⁰ Lambert and colleagues noted that there were no significant delays in the time to first treatment from diagnosis among patients with lung cancer treated during the pandemic in Manitoba.²¹ Finally, a 2024 study examined lung cancer diagnosis and management during the COVID-19 pandemic at a single thoracic surgical centre covering the province of Nova Scotia. In concordance with our study, the time to surgery was decreased in the postpandemic period; the authors found no differences in the number of referrals, operative approach, length of stay, or pathological stage. This group postulated that a pooled model of referrals was advantageous during this time in expediting patients’ surgical care, in contrast to surgeons receiving individualized referrals.²² Generally, these studies, as well as our own study, suggest that the timeline of surgical care for patients with lung cancer may have become more efficient during the pandemic. Potential mechanisms for this observation include altered referral patterns and prioritization of oncologic cases in a resource-constrained setting. Much of the relevant literature has also shown a trend toward “upstaging” of disease both during and after the pandemic.^4,20–22 It is too soon to understand whether the pandemic’s disruption to screening and early detection of lung cancer is transient, or if surgeons in Canada will continue to manage more advanced cancers going forward.

Some aspects of improved patient care described in our study that are byproducts of the pandemic can be applied to improving wait times outside of a global crisis. For example, during this time our group demonstrated that routine chest radiography after chest tube removal in the absence of clinical symptoms can safely be replaced with close clinical observation, which ultimately leads to significantly shorter length of stay and decreased health care costs and resource utilization.²³ Additionally, as was done in many other service-based industries, surgeons may have offered more virtual and telephone clinic visits to patients during the pandemic. This has not been studied in detail, but we postulate that this could facilitate higher volumes of patients being seen in consultation and triaged appropriately for surgery. This remains an ongoing area of research interest.

Limitations

The primary limitation of our study is the retrospective nature and inability to infer causality. Some variables of interest (pulmonary function testing, PET, pathological staging data) were not available for all included patients. Data on referral source were not consistently captured in our database to allow us to comment on lung cancer screening during the pandemic. Patients who received neoadjuvant therapy were excluded from this analysis, as their treatment pathway inherently includes longer planned intervals before surgery. Including this subgroup would have confounded the interpretation of system-related delays in surgical care. Finally, this study included only surgical patients, and we do not have data on the number of patients diagnosed with advanced, unresectable disease (owing to metastases or other factors) in the different cohorts.

Conclusion

In this single-centre, retrospective cohort study we describe a shorter time to surgery, shorter length of stay in hospital, and more advanced pathological disease among patients undergoing surgical resection of lung cancer during and after the COVID-19 pandemic than in preceding years. In the more recent years after the pandemic, wait times to surgery have increased to prepandemic levels, but patients continue to have more advanced disease at the time of resection. Our study suggests that, once patients were connected with a thoracic surgeon at our community hospital, timelines to surgical care were not delayed and may in fact have been expedited owing to an increase in dedicated resources during the pandemic. In synthesizing our findings of more advanced pathological disease at the time of resection with the available literature, we postulate that delays in initial presentation to a health care provider (i.e., screening or investigation of symptoms) may have contributed to this pattern. In the coming years, we may expect an increase in cases of advanced or nonoperative lung cancer, as well as related mortality. It is imperative to tailor the accessibility of diagnostic and treatment resources to patients early on to improve their access to care and cancer-related outcomes. Through further study of adaptive mechanisms such as these, we can continue to provide increasingly effective and efficient lung cancer care.