Lessons Learned: A Disruption in Care Leads to Increased Rates of Proximal Amputations

Beginning in the year 2020, the COVID-19 pandemic was responsible for a change in the way health care was administered in the United States. Hospital paradigms shifted to accommodate COVID-19 admissions. In an attempt to address the increased admission rate, hospital systems restricted elective surgeries and reduced the availability of health services deemed non-essential¹ Surgical procedures were prioritized where emergent surgery was needed. In addition, surgical care was reduced in patients with risk factors for limb threatening infections and chronic limb threatening ischemia (CLTI), such as diabetes mellitus (DM) and peripheral arterial disease (PAD). This is of particular concern in the patient population with wounds and infections secondary to these diseases, who would normally have access to treatment including wound care and debridement. Sallustro showed that non-healing vascular leg wounds experienced “an increase rate of superimposed infection, recurrence, and amputation rate” due to the restricted access patients had to wound care.² The potential delay in care put this vulnerable patient population at risk for eventual limb loss. 

Current rates of amputation in the United States have been increasing and are expected to increase to approximately 3.6 million people by the year 2050, with DM and PAD being the cause of 54% of the non-traumatic amputations.³ At present, there is a rise in diabetes with over 30 million Americans affected (8.3% of Americans). In a cohort of 962,496 foot ulcers from 2005-2010, diabetes was noted to account for 83 and 95% of major and minor amputations, respectively.⁴ Overall, the mortality rate associated with diabetic foot infections is found to be comparable to or greater than that of most types of cancers.⁵

PAD affects approximately 8.5 to 12 million Americans and has increased by about 25% over the last decade.⁶ In this period, vascular interventions have advanced, resulting in an overall decrease in amputation rates for CLTI.7 Despite these innovations, amputation rates secondary to PAD and CLTI are as high as 22%.⁷ This rate is attributed to their spectrum of severity in addition to their additional comorbidities. Within the substratum of PAD patients with CLTI, their prevalence of diabetes ranges from 27% to 76%.8 Their respective amputation risk is approximately four times higher, with several studies demonstrating that 25% to 90% of their amputations are attributable to their diabetic comorbidity.^8,9,10 

PAD in the presence of DM is known to exacerbate the development of the disease and leads to a poorer prognosis. It has been estimated that DM has been associated with a 2-to-4-fold increase in the prevalence and severity of PAD.¹¹ Compared to patients without diabetes, patients with DM have been shown to more frequently exhibit signs of infrapopliteal or tibial artery disease.¹² The increased risk of amputations in patients with DM and PAD concomitantly is thought to be associated with the diabetic impaction of autonomic neuropathy leading to resultant foot deformities along with impaired wound healing in the presence of impaired arterial flow.¹³ 

This study aimed to determine if there was a change in the rate of proximal amputations, including below knee and above knee amputations, performed during the COVID-19 pandemic. Because of the decreased patient follow up, decreased non-emergent limb salvage procedures, and limited availability of peripheral intervention on patients with diabetes and CLTI, we hypothesized that the rate of amputations during the mandated shutdown period of the pandemic will increase.

This was a retrospective study performed within the Northwell Health Hospital System, a quaternary care system located in the greater New York area, which includes 23 hospitals in Brooklyn, Queens, Staten Island, Manhattan, Westchester County, and Long Island. Institutional Review Board approval was obtained under the Northwell Health COVID-19 Research Consortium (IRB # 20-0200-OTH). 

Outcome Measures
The primary outcome measures were amputations performed between December 2019 and September 2020. Amputation types were one of two categories: proximal amputations (above knee, below knee amputations), or pedal amputations (trans-metatarsal amputations, partial ray amputations, and digit amputations). The inclusion criteria were patients older than 18 years old with the diagnosis of peripheral artery disease, diabetes mellitus, or a combination of those diseases.

Database/Patient Cohort
Using the Sunrise electronic medical record, data regarding system-wide amputations were obtained from the pre-specified dates: January 7, 2020 through March 19, 2020 (pre-pandemic), March 20, 2020 through May 31, 2020 (shutdown), June 1, 2020 through August 12, 2020 (post-shutdown), and August 13, 2020 through January 13, 2021 (reopening). 

The pre-pandemic period was determined to be 73 days before the state-wide lockdowns were enforced in New York. The shutdown time frame was designated as the 73 days that the lockdowns were strictly in effect and in-hospital restrictions were placed on non-urgent and emergent surgery. The post-shutdown period was allocated to the 73 days when a tiered surgical structure prioritized more urgent surgical cases to be performed. The reopening period lifted the restrictions on non-urgent and emergent surgeries.

Patient demographics (age, sex, race) and medical history including comorbidities (diabetes, tobacco use, PAD, hyperlipidemia, coronary artery disease, stroke, cancer, and chronic obstructive pulmonary disease) were analyzed. Amputations were then categorized as infectious, ischemic, or both, utilizing electronic medical records (Allscripts Sunrise).

Ischemic PAD was designated to any patients with an ankle-brachial index (ABI) < 0.9. If ABI > 1.3 or non-compressible vessels, toe brachial index data was evaluated. Patients with a toe-brachial index (TBI) value < 0.7 were noted to have PAD etiology.

Amputations performed on patients with normal vascular testing were categorized as infectious etiology. Amputations on patients with an ABI < 0.9/TBI < 0.7 with infection present in the wound were categorized as both infectious and ischemic etiology.

Statistical Analysis
Descriptive statistics were used to describe the overall study sample, as well as stratified by time period and type of amputation. Patient age and length of stay were the only two continuous variables. Age was normally distributed and thus, means and standard deviations were reported, whereas length of stay was not normally distributed and thus, medians and interquartile ranges (IQRs) were reported. Categories' variables are reported and frequencies and proportions. Differences in patient characteristics by time period and type of amputations were assessed using analysis of variance (ANOVA), as well as t, Kruskal-Wallis, and Wilcoxon rank sum tests for continuous variables, and chi-square and Fisher’s exact tests for categorical variables.  The proportion of proximal amputations for each time period was reported and a chi-square test was used to assess differences in the proportions across the four time periods. A P value of < 0.05 was considered statistically significant.

A total of 579 patients were included in the final analysis. There were no significant differences in patient demographics including age, sex, race, and ethnicity between the periods (Table 1). There were more patients with stroke and cancer and use of non-steroidal anti-inflammatory drugs (NSAIDs) in the reopening time period compared with other time periods. As shown in Table 2, when stratified by type of amputation, patients undergoing proximal amputations were generally older compared with those undergoing pedal amputations (mean age of 70 vs. 64 years; P=0.0002). In addition, patients undergoing proximal amputations were found to have higher rates of PAD (45.4% vs 28%), stroke (19.8% vs 9.7%), and prior history of vascular surgeries (45.4% vs 34.3%). Diabetes was more prevalent in the patients who underwent pedal amputations (87.4% vs 65.1%). This was also reflected in the finding that pedal amputations were performed more often for infection (61.1% vs 31.4%), whereas proximal amputations were performed more often for PAD (30.2% vs 7.5%). 

The pre-pandemic period had a total of 179 amputations performed, with 8.38% of the amputations being proximal (Table 2). In comparison, during the shutdown period, there were only 86 amputations performed, with a greater proportion of them being proximal amputations (25.58%, P=0.0009). Following the shutdown period, the number of amputations returned to baseline and the proportion of proximal amputations decreased throughout the post-shutdown and reopening periods (18.5% and 12.06%). This was equivalent to significantly  higher odds of undergoing a proximal amputation if performed during the Shutdown (4.89 OR, Table 3). The post-shutdown period also had a significantly higher odds of proximal amputation (3.12 OR); however, this difference was lost in the reopening period (1.98 OR).

The COVID-19 pandemic shifted the way health care was performed as alternative ways to deliver care were sought out to reduce transmission rates. In the United States, mandatory shutdowns and limited in-person visits had effects on those patients with risk factors for limb loss, including diabetes and peripheral vascular disease. There are currently limited studies published evaluating amputation risk during the pandemic. To date, this is the only study that evaluated types of amputations and their etiologies. 

This study shows that the proportion of proximal limb loss was significantly higher during the shutdown period of the pandemic. We suspect that the increase in proximal amputations during this time frame is due to the inadequate access to necessary healthcare in addition to the patients’ fear of contracting COVID-19, preventing them from seeking appropriate treatment. The patient’s lack or avoidance of treatment or distal amputations led to a large portion of poorly supervised wounds and infections, which resulted in deterioration, ultimately necessitating major amputations that would have been otherwise unnecessary. 

Similar results for higher rates of amputations have been found in other areas of the United States. Casciato et al demonstrated an abrupt decrease in hospital admissions after lockdown, but a statistically significant increase of emergent surgeries was seen during the pandemic period.¹⁴ They also found that the odds of an amputation at any level during the pandemic period was 10.8 times higher than the pre-pandemic period and the odds of a major amputation during the pandemic period was 12.5 times higher when compared to the pre-pandemic period. Lancaster et al also observed an increase in the major amputation rate during the pandemic period when compared to pre-pandemic levels.¹⁵ They attributed the increased rate of amputation to the delayed presentation for patients and limited surveillance mechanisms. 

These findings are replicated in other areas of the world affected by COVID-19. In the Netherlands, similar lockdown effects took place, with restricted outpatient and primary care physician visits, and new consultations performed via phone. Schuivens et al found that more major amputations were performed during the pandemic in 2020, with a rate of 42% compared to 18% in the years prior.¹⁶ They also showed that during the pandemic, patients presented with higher grades of CLTI consistent with more extensive ischemia.

However, in Poland a significant rise in minor amputations was noted (from 3146 to 4269) due to an increase in urgent hospital admissions amongst diabetic foot ulcers (DFU) patients.¹⁷

Italy had one of the highest rates of death and overall COVID-19 transmission, leading to an essentially complete lockdown of the country. Caruso et al showed that there was a significantly higher rate of emergency admission for diabetic foot ulcers during this period, as well as a higher proportion of patients requiring amputation (60% vs 18%). This equated to an approximate threefold risk of amputation compared to the year prior.¹⁸ Stabile et al, in Campania, Italy, noted that there was a statistically significant increase in amputations, with those patients being admitted already suffering from advanced tissue loss during the lockdown period.¹⁹

In Mexico City, Lozano-Corona et al had similar findings with a “two-fold increase in major amputations in the COVID-19 era (41% vs 21%).” It was also noted that the number of revascularization procedures performed for limb salvage significantly decreased and major amputations were occurring in mainly non-revascularized patients (37% vs 17%).²⁰ Lou et al studied the impact of vascular procedures in North America and found that claudication procedures decreased by about 70% compared to before lockdown limitations were placed.²¹ The inability to perform preventative treatment on patients further reinforces the diminishing impact that COVID-19 had on healthcare and treatment options. 

Fortunately, as the period extends further out from the pandemic and health care systems resume routine care, we start to see the effects of the pandemic diminish as the proportion of proximal amputations decreases and becomes more comparable to that of before the pandemic (Table 2). This is encouraging in that even as changes in rules and regulations have remained, the adjustments made by health care systems to provide continued care are working. This stresses the importance and necessity of availability of continued wound care treatments. Such treatments are needed for these patients with an elevated risk of limb loss. As COVID-19 variants persist and the threat of resuming lockdowns rises once more, the fact remains that these patients must have ongoing care provided. 

The major strength of this study was that this study had the largest patient population studied to date in a region that was significantly affected by the pandemic. Additionally, the extended timeline incorporated in this study allows for the analysis of the health care system’s response to the pandemic. Limitations include findings within a single health care system in one region of the United States. Other regions may have different resource utilization and different responses to the pandemic in treating this patient population.

There are several limitations to this study. Missed follow-up appointments with vascular and/or diabetes specialists and the inability to refill prescriptions could also possibly affect the outcome of increased amputations. Additionally, a lack of close monitoring during the initial months of the pandemic could have influenced the progression of poor outcomes.

The increased proportion of limb loss as a result of the COVID-19 pandemic highlights the importance of routine follow-ups for at-risk populations. The role of continued wound care, non-emergent vascular procedures, and toe and forefoot amputations is critical in preventing proximal amputations. The pandemic illustrated the acuity of patients who are at increased risk of amputation should a major disruption to access to health care recur. The fact that a majority of those amputations were proximal amputations, with an increased impact on the patient, reinforces the need for providers to be cognizant of this vulnerable population. Future disaster management plans should include a means of access to ensure that these individuals are not overlooked. In the event elective surgeries were to be canceled again, we recommend that an exception is made to this subset of patients to prevent future limb loss.

¹Resident Physician, Department of Surgery, Zucker School of Medicine at Hoftsra/Northwell, Hempstead, NY
²Department of Surgery, Comprehensive Wound Healing & Hyperbaric Center, Northwell Health, Lake Success, NY
³Resident Physician, Department of Podiatry, Long Island Jewish/North Shore, New Hyde Park, NY
⁴Department of Podiatry, North Shore University Hospital Specialty Clinic, Northwell Health, Manhasset, NY
⁵Department of Emergency Medicine, Zucker School of Medicine at Hoftsra/Northwell, Hempstead, NY
⁶Division of Vascular and Endovascular Surgery, Zucker School of Medicine at Hofstra/Northwell, Hempstead, NY

Address all correspondence to: Alisha R. Oropallo, MD, Northwell Health, 1999 Marcus Avenue, Suite M6, Lake Success, New York 11042, Tel: 516-233-3780, Fax: 516-233-3644, Email: aoropallo@northwell.edu
 

none disclosed

none