If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Telephone visits had a higher proportion of minority and underinsured patients.
•
No difference in 30-day hospital encounters for video vs. telephone visits.
•
Lower 30-day hospital encounter rate for pandemic vs. historical video visits.
•
Patients who did not meet study inclusion criteria did not have worse outcomes.
Abstract
Background
Due to the COVID-19 pandemic, post-discharge virtual visits transitioned from a novel intervention to standard practice. Our aim was to evaluate participation in and outcomes of virtual post-discharge visits in the early-pandemic timeframe.
Methods
Pandemic cohort patients were compared to historical patients. Patient demographics, clinical information, and post-discharge 30-day hospital encounters were compared between groups.
Results
The historical cohort included 563 patients and the pandemic cohort had 823 patients. There was no difference in 30-day hospital encounters between patients who completed a video vs. telephone visit in the pandemic cohort (3.8% vs. 7.6%, p = 0.11). There was a lower 30-day hospital encounter rate in pandemic video and telephone visits compared to similar historical sub-groups.
Conclusion
Expansion of virtual post-discharge visits to include all patients and telephone calls did not negatively impact rates of 30-day post-discharge hospital encounters. Offering telehealth options for post-discharge follow-up does not appear to have negative impact on healthcare utilization.
Prior to the SARS-CoV-2 (COVID-19) pandemic, utilization of telehealth to augment traditional surgical care was a relatively novel concept that was being implemented and evaluated sporadically across the United States.
Due to the novelty of such interventions and overall low utilization, studies were primarily focused on feasibility and patient satisfaction.
Prior to the pandemic our surgical group ran a non-inferiority randomized control trial comparing post-discharge video-based virtual visits for patients who had undergone a laparoscopic appendectomy or laparoscopic cholecystectomy.
When the COVID-19 pandemic occurred, the research study was closed as it was no longer safe to randomize patients to an in-person visit due to the risk of viral spread and in-person contact. Simultaneously, we pivoted to providing virtual care (expanding to include telephone visits) as the gold standard and were able to do so quickly due to having infrastructure built to provide care virtually for patients. Notably, patients that had previously been ineligible to participate in a virtual visit were now - by default - offered a virtual visit.
The aim of our study was to examine similar patient cohorts in the pre-COVID-19 era and the post-COVID-19 era to better understand access, barriers, and outcomes to utilization of virtual visits across both timeframes. We hypothesized that the inclusion criteria for the randomized control trial, though intended to be pragmatic, was overly stringent and that patients who would have previously been excluded who completed a virtual visit would have similar outcomes to the original study intervention population.
2. Methods
Two cohorts of patients were compared: 1. Historical Cohort – patients from our historical data during the time period we were enrolling patients in the randomized control trial (August 2017–March 2020) and 2. Pandemic Cohort - patients in the COVID pandemic time-period (March 2020–April 2021). To examine different comparison cohorts within the historical vs pandemic time-period, subjects were divided into four sub-groups for comparison as shown in Fig. 1.
In both the Historical and Pandemic cohorts, patients aged 18–90 were identified based on having a minimally invasive (laparoscopic or robotic) appendectomy or cholecystectomy by a group of surgeons who cover non-assigned emergency general surgery (EGS) call at nearby hospitals (a tertiary care Level 1 trauma center with >850 beds and a 196-bed community hospital, with an additional Level 3 trauma center in the Pandemic cohort) in an urban setting. In the historical cohort, patients were eligible for recruitment if they spoke English, had a North Carolina or South Carolina residence, and had an email address and the technology required to complete a visit (Fig. 2). Patients were excluded from the historical study if they had perforated appendicitis or perforated cholecystitis, active cocaine use, or took narcotics for chronic pain.
Fig. 2Eligibility for virtual visits in the Historical and Pandemic cohorts.
In response, our clinic transitioned practice from enrolling patients in a randomized control trial with postoperative virtual visits as the intervention arm being studied to evaluate non-inferiority to making virtual visits standard. This response was necessary to reduce the risk of COVID-19 transmission and because it no longer remained ethical to randomize patients to in-person visits due to the risk of viral transmission. Patients in the pandemic cohort were instructed to follow up with a virtual visit whenever possible. Prior to availability of interpreting services, patients were offered a video-based virtual visit if they could speak English and had the required technology. Patients who did not meet these two criteria were offered a post-operative telephone-based visit. Patients who desired an in-person visit, were unable to complete a video or telephone virtual visit, or who had postoperative complications were scheduled for in-person follow-up. Patients who presented with perforated appendicitis or cholecystitis were not excluded from virtual visits in the pandemic cohort. In both cohorts, patients with drains or discharge to location other than home were excluded. When discussing with patients follow-up options, for virtual visits we specifically asked if there were family members or other care partners who could assist in completing the visit before offering telephone or in-person visits.
Virtual visits were completed by either an advanced practice provider (APP) or a surgeon- “surgical team member” will be used to refer to both roles. In February 2021 we were able to include non-English speaking patients in video-based visits due to a change in the video platform and integration of interpreting services (Fig. 3). Patient demographics, virtual visit type (video or telephone), and attendance were obtained from the electronic medical record (EMR). A priori identified comorbidities (diabetes, hypertension, congestive heart failure, bleeding disorders, anticoagulation) were identified and recorded by the research team. Due to small numbers of each individual comorbidity, presence of any comorbidity is reported in tables.
Our primary outcome of interest was postoperative hospital encounter utilization rates (emergency department, observation, or inpatient admission) within 30 days after surgery. Secondary outcomes were patient demographics, length of stay, timing of surgery (elective vs. emergency) and type of surgery (appendectomy vs. cholecystectomy). Specific postop complications other than utilization of healthcare were not identified and the type of post-discharge visit was not anticipated to impact complication rates. During the historical cohort any adverse events were reviewed to determine if they were related to the virtual follow-up to ensure safety of the randomized control trial.
Summary statistics for both continuous (mean ± standard deviation) and categorical variables (count and percentage) were compared for all groups. Continuous data was compared with t-test or Mann-Whitney. Categorical data was compared using the Pearson's Chi-squared test or Fisher's exact test, as appropriate. All tests will be two-sided with a significance level of 0.05. Statistical analysis was performed using SAS Enterprise Guide version 7.15 (SAS Institute, Cary, NC). This study was approved by the Atrium Health Institutional Review Board (IRB).
3. Results
We identified 563 patients from our Historical Cohort and 823 patients from the Pandemic Cohort. During the Pandemic time period, 19% of patients had a video visit and 33% of patients completed a telephone visit. Video visit patients had a higher percentage of white patients and patients with either Medicare or private insurance (Table 1). Telephone visit patients had a higher percentage of black patients and patients who were either un-insured or underinsured. Patients reporting Hispanic ethnicity were more likely to request a telephone visit, which was the only virtual option available for Spanish-speaking patients for the majority of the pandemic time. A higher percentage of patients who had video visits had an appendectomy performed. There was no difference in post-discharge care utilization (30-day hospital encounter) between video and telephone visits in the Pandemic Cohort.
Table 1Pandemic video vs phone group characteristics and outcomes.
Patient Characteristics
Video visits N = 160)
Phone visits (N = 275)
p-value
Travel minutes*
24.4 ± 30.6
25.7 ± 46.9
0.75a
Travel miles*
15.0 ± 28.6
17.4 ± 47.9
0.57a
Age category, years
0.18c
18-30
46 (28.8)
82 (29.8)
30-40
38 (23.8)
59 (21.5)
40-50
33 (20.6)
40 (14.5)
50-60
15 (9.4)
43 (15.6)
60-70
19 (11.9)
26 (9.5)
>70
9 (5.6)
25 (9.1)
Female sex
88 (55.0)
168 (61.1)
0.21c
Race
<0.001c
White
122 (76.3)
140 (50.9)
Black
31 (19.4)
77 (28.0)
Other
5 (3.1)
16 (5.8)
Unknown
2 (1.3)
42 (15.3)
Insurance
<0.001c
Commercial, medicare, managed care
137 (85.6)
159 (57.8)
Medicaid, selfpay, workers comp, other
23 (14.4)
116 (42.2)
Ethnicity
<0.001d
Hispanic or Latino
12 (7.5)
81 (29.5)
Not Hispanic or Latino
144 (90.0)
191 (69.5)
BMI
0.50d
BMI <25
36 (22.5)
54 (19.6)
BMI 25.0–29.9
52 (32.5)
77 (28.0)
BMI 30–34.9
36 (22.5)
65 (23.6)
BMI ≥ 35
34 (21.3)
70 (25.5)
Unknown
2 (1.3)
9 (3.3)
Any comorbidity (Yes)
86(53.8)
131(47.6)
0.22c
Patient Outcomes
Length of stay, days
0.20c
<1
66 (41.3)
110 (40.0)
= 1
78 (48.8)
121 (44.0)
>1
16 (10.0)
44 (16.0)
Surgery Type
0.007c
MIS appendectomy
89 (55.6)
116 (42.2)
MIS cholecystectomy
71 (44.4)
159 (57.8)
Timing of Surgery
0.35c
Scheduled
7 (4.4)
18 (6.5)
Unplanned
153 (95.6)
257 (93.5)
30-day Hospital encounter
6 (3.8)
21 (7.6)
0.11c
30-day inpatient Readmission
2 (1.3)
8 (2.9)
0.34d
30-day observation readmission
0 (0.0)
3 (1.1)
0.30d
30-day ED encounter
5 (3.1)
14 (5.1)
0.33c
*Data not available for all subjects. Values presented as Mean ± SD, Median [P25, P75], Median (min, max) or N (column %). p-values: a = ANOVA, b = Kruskal-Wallis test, c = Pearson's chi-square test, d = Fisher's Exact test. MIS = minimally invasive surgery.
When comparing patients who would have been study-eligible in the pandemic video visit cohort to those in the historical video visit cohort, we found that a higher percentage of patients were of white race during the pandemic time-period compared with historical data (Table 2). The pandemic patient cohort had a higher proportion of patients with shorter lengths of stay (≤1 day). During the pandemic time-period, a smaller percentage of video visit patients had elective surgery. Also noted during the pandemic time-period, patients had a lower percentage of 30-day hospital encounters and 30-day observation readmissions.
Table 2Comparison A: Comparing Pandemic video visits (study-eligible) vs. Historical video visits.
Pandemic video visits (N = 140)
Historical video visits (N = 135)
p-value
Age category, years
0.19d
18-30
42 (30.0)
42 (31.1)
30-40
33 (23.6)
36 (26.7)
40-50
27 (19.3)
31 (23.0)
50-60
14 (10.0)
17 (12.6)
60-70
17 (12.1)
6 (4.4)
>70
7 (5.0)
3 (2.2)
Female sex
80 (57.1)
85 (63.0)
0.32c
Race
0.002d
White
109 (77.9)
91 (67.4)
Black
27 (19.3)
27 (20.0)
Other
4 (2.9)
6 (4.4)
Unknown
0 (0.0)
11 (8.1)
Insurance
0.088c
Commercial, Medicare, managed care
120(85.7)
105(77.8)
Medicaid, self pay, workers comp, other
20(14.3)
30(22.2)
Ethnicity
0.83d
Hispanic or Latino
9(6.4)
9(6.7)
Not Hispanic or Latino
127(90.7)
120(88.9)
BMI (kg/m2)
0.19d
BMI <25
32(22.9)
42(31.1)
BMI 25.0–29.9
45(32.1)
44(32.6)
BMI 30–34.9
28(20.0)
14(10.4)
BMI ≥35
33(23.6)
33(24.4)
Unknown
2(1.4)
2(1.5)
Any comorbidity (Yes)
75(53.6)
60(44.4)
0.13c
Length of stay (days)
0.003c
<1
65(46.4)
39(28.9)
= 1
68(48.6)
78(57.8)
>1
7(5.0)
18(13.3)
Surgery Type
0.23d
Laparoscopic appendectomy
73(52.1)
60(44.4)
Laparoscopic or robotic cholecystectomy
67(47.9)
75 (55.5)
Timing of Surgery – scheduled (elective)
6(4.3)
19(14.1)
0.005c
30-day Hospital encounter
6 (4.3)
15 (11.1)
0.033c
30-day inpatient Readmission
2 (1.4)
3 (2.2)
0.68d
30-day observation readmission
0 (0.0)
5 (3.7)
0.027d
30-day ED encounter
5 (3.6)
8 (5.9)
0.36c
Values presented as Mean ± SD, Median [P25, P75], Median (min, max) or N (column %).
p-values: a = ANOVA, b = Kruskal-Wallis test, c = Pearson's chi-square test, d = Fisher's Exact test.
Patients who were study-ineligible due to inability to perform a video-based visit were compared to pandemic patients who completed a telephone visit to examine if a telephone visit would be adequate for patients that cannot do a video-based virtual visit. During the pandemic time-period, a higher percentage of patients who did telephone visits were white and over seventy years of age, and a higher percentage of them had at least one identified comorbidity (detail on comorbidities presented in Supplemental Table 1). While laparoscopic cholecystectomy was the most common procedure in both groups, in the pandemic time period a much higher percentage of telephone visits were after a laparoscopic appendectomy. Pandemic period patients who performed telephone visits had a shorter length of stay and a lower 30-day hospital encounter and 30-day emergency department encounter rate than historical patients unable to complete a video visit and assigned to an in-person post-discharge visit. Also noted a higher percentage of patients with less than one day length of stay. The historical study-ineligible cohort had a higher percentage of patients of Hispanic ethnicity (Table 3). In additional analysis adjusting for length of stay and elective vs. unplanned surgery (as markers of severity of illness) and comorbidities in multivariable regression, care utilization was not increased in the setting of expanded virtual visits in the pandemic cohort (OR = 0.55, 95%CI: 0.31–0.7).
Table 3Comparison B: Characteristics and outcomes of patients unable to complete a video-based visit.
Factor
Pandemic telephone visits (N = 275)
Historical study-ineligible patients (N = 327)
p-value
Age category, years
0.011c
18-30
82 (29.8)
87 (26.6)
30-40
59 (21.5)
108 (33.0)
40-50
40 (14.5)
49 (15.0)
50-60
43 (15.6)
36 (11.0)
60-70
26 (9.5)
32 (9.8)
>70
25 (9.1)
15 (4.6)
Female sex
168 (61.1)
229 (70.0)
0.021c
Race
<0.001c
White
140 (50.9)
63 (19.3)
Black
77 (28.0)
38 (11.6)
Other
16 (5.8)
25 (7.6)
Unknown
42 (15.3)
201 (61.5)
Insurance
<0.001c
Commercial, Medicare, managed care
159 (57.8)
64(19.6)
Medicaid, self pay, workers comp, other
116 (42.2)
263(80.4)
Ethnicity
<0.001c
Hispanic or Latino
81(29.5)
201(61.5)
Not Hispanic or Latino
191(69.5)
114(34.9)
BMI (kg/m2)
0.11c
BMI <25
54(19.6)
44(13.5)
BMI 25.0–29.9
77(28.0)
112(34.3)
BMI 30–34.9
65(23.6)
86(26.3)
BMI ≥35
70(25.5)
70(21.4)
Any comorbidity (Yes)
131(47.6)
106(32.4)
<0.001c
Length of stay (days)
<1
110(40.0)
80(24.5)
<0.001c
= 1
121(44.0)
196(59.9)
>1
44(16.0)
51(15.6)
Surgery Type
Laparoscopic appendectomy
116(42.2)
87(26.6)
<0.001c
Laparoscopic cholecystectomy
159(57.8)
240(73.4)
Timing of Surgery – scheduled (elective)
18(6.5)
52(15.9)
<0.001c
30-day hospital encounter
21(7.6)
43(13.1)
0.029c
30-day inpatient Readmission
8(2.9)
12(3.7)
0.60c
30-day observation readmission
3(1.1)
6(1.8)
0.52d
30-day ED encounter
14(5.1)
34(10.4)
0.017c
Values presented as Mean ± SD, Median [P25, P75], Median (min, max) or N (column %).
p-values: a = ANOVA, b = Kruskal-Wallis test, c = Pearson's chi-square test, d = Fisher's Exact test.
During both the historical and pandemic time periods, there were patients who did not complete a post-discharge visit - both for in-person visits as well as telemedicine visits (47% Pandemic and 18% Historical). We compared patients who were scheduled for a video-based virtual visit and did not attend in both our historical and pandemic data (Table 4). During the pandemic, older patients (50–60, 60–70 and > 70 years) made up a higher percentage of patients not attending their visits. In the historical cohort, black patients had a higher rate of non-attendance compared to the pandemic time period.
Table 4Comparison C: Patients without a postoperative clinic visit in 30 days.
Factor
Pandemic video visit patients who did not f/u (N = 388)
Historical video visit patients who did not f/u (N = 101)
In Table 5, we compare pandemic video visit study-eligible to pandemic video visit study-ineligible participants. During the pandemic, higher-risk patients who would have been excluded from the randomized control trial were offered video-based visits due to a new risk/benefit balance. The pandemic study-ineligible participants had a lower percentage of white patients. This group was also noted to have a higher percentage of patients with a length of stay greater than 1 day. There were no differences in outcomes between these groups.
Table 5Comparison D: Comparing study-eligible and study-ineligible Pandemic video-based visits.
Factor
Pandemic period study-eligible (N = 140)
Pandemic period study-ineligible (N = 20)
p-value
Age category, years
0.70d
18-30
42 (30.0)
4 (20.0)
30-40
33 (23.6)
5 (25.0)
40-50
27 (19.3)
6 (30.0)
50-60
14 (10.0)
1 (5.0)
60-70
17 (12.1)
2 (10.0)
>70
7 (5.0)
2 (10.0)
Female sex
80 (57.1)
8 (40.0)
0.15c
Race
0.021d
White
109 (77.9)
13 (65.0)
Black
27 (19.3)
4 (20.0)
Other
4 (2.9)
1 (5.0)
Unknown
0 (0.0)
2 (10.0)
Insurance
0.99d
Commercial, Medicare, managed care
120 (85.7)
17 (85.0)
Medicaid, self pay, workers comp, other
20 (14.3)
3 (15.0)
Ethnicity
0.31d
Hispanic or Latino
9 (6.4)
3 (15.0)
Not Hispanic or Latino
127 (90.7)
17 (85.0)
BMI (kg/m2)
0.17d
BMI <25
32 (22.9)
4 (20.0)
BMI 25.0–29.9
45 (32.1)
7 (35.0)
BMI 30–34.9
28 (20.0)
8 (40.0)
BMI ≥35
33 (23.6)
1 (5.0)
Any comorbidity (Yes)
75(53.6)
11(55.0)
0.90c
Length of stay (days)
<0.001d
<1
65 (46.4)
1 (5.0)
= 1
68 (48.6)
10 (50.0)
>1
7 (5.0)
9 (45.0)
Surgery Type
0.019c
Laparoscopic appendectomy
73 (52.1)
16 (80.0)
Laparoscopic cholecystectomy
67 (47.9)
4 (20.0)
Timing of Surgery – scheduled (elective)
6 (4.3)
1 (5.0)
0.99d
30-day hospital encounter
6 (4.3)
0 (0.0)
0.99d
30-day inpatient Readmission
2 (1.4)
0 (0.0)
0.99d
30-day observation readmission
0 (0.0)
0 (0.0)
30-day ED encounter
5(3.6)
0(0.0)
0.99d
Values presented as Mean ± SD, Median [P25, P75], Median (min, max) or N (column %).
p-values: a = ANOVA, b = Kruskal-Wallis test, c = Pearson's chi-square test, d = Fisher's Exact test.
This study evaluated utilization and outcomes of virtual post-discharge surgical care during the COVID-19 pandemic and in comparison, to historical cohorts of patients screened for enrollment in a randomized control trial. We found that in the pandemic time period, patients undergoing telephone calls over video visits were more likely to be of a minority race, Hispanic ethnicity, and underinsured, but there was no statistical difference in 30-day hospital use between the two groups. While not significant, the rate of 30-day hospital use for patients who had a telephone call (7.6%) was noted to be twice that of those who had a video visit (3.8%). Both rates remain lower than historical cohorts, it would be important to monitor them going forward for clinics that continue to provide both telephone and video-based options. Our primary outcome of 30-day hospital encounter included inpatient readmissions, observation readmissions and any ED encounter. The rate of 30-day inpatient readmission (1.3% and 2.9%) are also lower than the historical cohorts and the same or lower as those reported in the literature (2.0–5.9%).
Lower readmissions during the pandemic time period were likely influenced by a higher threshold for admission from the healthcare team's perspective as well as patients' desire to avoid healthcare interactions to decrease risk of exposure to COVID. Interestingly, this is opposite of our findings in a previous evaluation of a broader of cohort of patients with an emergency general surgery diagnoses across our regional system, where readmissions were higher in the pandemic period compared to a historical cohort.
This may highlight differences in low-risk and high-risk EGS diagnoses, as well a difference in readmission for patients managed operatively versus non-operatively.
Prior to the COVID-19 pandemic, telemedicine was underutilized in the general surgery population. Telemedicine was predominantly used to either triage patients for transfer to higher level of care or to seek consultation from a specialist or colleague. Most of these methods were utilizing a store-and-forward format – an asynchronous technique – meaning the picture would be taken and sent for review at the receiving institutions earliest convenience. In the years leading up to the COVID-19 pandemic our group sought to provide synchronous telemedicine visits utilizing video-based visits for our post-surgical patients and evaluate outcomes through a randomized control trial.
When the COVID-19 pandemic arrived, the focus shifted to providing virtual care to as many patients as possible utilizing both video-based and telephone-based telemedicine venues to decrease the spread of COVID-19. However, this provided a unique ability to examine patient demographics and outcomes during the pandemic period to our historical data. During the pandemic time period, video visit patients and patients not able to complete a video visit had a lower incidence of 30-day hospital encounter compared to their historical cohorts. There was no meaningful difference between outcomes for pandemic video visit patients (study eligible vs. study ineligible) or patients who did not complete a post-discharge visit in 30 days (pandemic time period vs. historical).
One aspect of virtual care that was novel in the pandemic time period was the introduction of telephone-based post-discharge virtual visits. We had previously believed that the ability to examine wounds and see patients would be an important part of post-discharge assessment. However, successful implementation of post-discharge telephone visits had been previously reported.
Telephone-based virtual visits were incorporated in our clinic during the pandemic, and we found no difference in outcomes between video-based and telephone-based visits, while also finding that minority race, Hispanic ethnicity, and under- and uninsured patients were more likely to choose the telephone format. This raises important questions about disparities in access to virtual care, particularly as we transition to a post-pandemic time period when clinical and policy decisions are being made about which modes of virtual care to continue and how they should be reimbursed.
on patients who do not have access to that type of care?
When we examined outcomes of video-visit patients between those who participated in the historical research study and those who met study criteria during the pandemic time period, the most notable differences were an increase in non-elective surgery, shorter length of stay, and decreased 30-day hospital encounters in the pandemic period. These are all anticipated outcomes given clinical changes made to adapt to providing care in the pandemic period.
With high volumes of patients hospitalized, there was a greater need for patients to be discharged home in an expedited fashion resulting in decreased length of stay in the pandemic. Notably, this finding may be specific to patients undergoing low-risk procedures, as prior research has demonstrated no difference in length of stay for emergency general surgery patients in the pandemic vs. pre-pandemic times.
Many patients were also motivated to minimize the amount of time they were hospitalized, as to decrease their exposure to COVID-19. In our practice, most patients were discharged home either the same day as their procedure or the day after. While decreased utilization of hospital care during the pandemic has been demonstrated,
it is notable that even for patients who did receive hospital care, they were less likely to have a post-discharge 30-day hospital encounter. This may signify the beneficial impact of efforts to increase resources to support patients in the outpatient setting.
Pandemic cohort patients who would have been historical study-ineligible were offered a telephone visit for post-discharge surgical care. When comparing their outcomes to historical study-ineligible patients the pandemic patient population had a higher percentage of patients over the age of 70, a higher percentage of non-Hispanic patients and a higher percentage of patients with commercial/Medicare/managed care insurance who had follow-up telephone visits, potentially indicating that this option had a broad appeal to patients who would have previously been limited to either a video-based or in-person follow-up option. Since the pandemic cohort patients had shorter length of stay and lower 30-day hospital encounter and ED encounter rates, it would suggest that telephone visits are not a risker option than an in-person post-discharge visit. Similar to the comparison between pandemic and historical video visits, this comparison is impacted by notably different care paradigms between the two time periods.
Postoperative visits are important for patients for several reasons: to assess their overall recovery status, discuss important ongoing restrictions, address health maintenance, and review pathology and any next steps required (such as in the case of unexpected malignancy, which is fortunately very rare).
Unfortunately, some patients may feel that the postoperative visit is not an important step in their surgical care, especially if they are feeling well at that time. Patients were asked prior to arranging a virtual visit if they had the appropriate equipment such as a smartphone to be able to perform the visit, and many patients were working from home during the pandemic. Travel time and mileage is known to be associated with decreased rates of completing a post-discharge visit
and these potential barriers are removed for virtual visits, making it notable that there remain other barriers to completing post-discharge visits. Future research should continue to investigate reasons why patients do not attend post-discharge surgical visits and identify methods to address patient needs. If visits are inconvenient from a timing perspective, maybe a store-and-forward communication method would be sufficient.
Alternatively, if the visits are perceived to be low value, understanding ways to increase the value may be beneficial.
Though the study design for the historical randomized control trial was intended to be pragmatic and inclusive, there were exclusion criteria to rule out the patients at highest risk for postoperative complications – notably perforated appendicitis and a length of stay >4 days. In the pandemic time period, these patients were offered a video visit due to changes in the risk/benefit balance in the setting of potential for COVID-19 transmission. This previously identified “high-risk population” had a longer length of stay and higher percentage of patients who underwent appendectomy, both of which were anticipated based on the exclusion criteria. Although no significant difference was found when we compared the outcomes of these patients to the study-eligible patients who completed video visits, the patient numbers were small and comparisons were likely under-powered. Despite this, it is likely feasible to offer virtual post-discharge visits to a broader population of postoperative patients when a robust in-person safety net clinic is available.
4.1 Limitations of study
One limitation of the study was that there was no ability to provide non-English speaking patients the option of video virtual visits until February 2021, which was well into the COVID-19 pandemic. Multiple attempts were made pre-pandemic to be able to provide multi-lingual visits however this was a limitation of both the platform at the time and resource availability of interpreters. This would be something to study in the future, as discussions with our Hispanic Hospital Liaison discussed that although the majority of Spanish-speaking patients do have smart-phones they utilize them in different ways than their English-speaking counterparts. Additionally, we lack detailed information on factors that may have influenced patients to choose telephone follow-up, which may also influence rates of post-discharge hospital care utilization.
All comparisons may be limited by low sample size due to limitations of a cohort study. In particular, when comparing the pandemic study-eligible and pandemic study-in-eligible outcomes, this data is limited due to the low number in the ineligible cohort. Though an overall 563 patients in historical cohort and 823 patients in pandemic cohort would achieve a power of 0.84 in detecting a 5% difference, the power of our predetermined subgroup comparisons runs lower than the typical power of 0.8. Additional studies with a larger sample size if feasible would be valuable to confirm our findings.
During the pandemic, higher overall patient census and decreased ability to schedule elective cases could have differentially impacted our outcomes. Although our data captured encounters at any hospital or freestanding ED within our regional health system, it remains possible that patients presented to other hospitals that were not captured within our data warehouse and may not have been captured.
5. Conclusion
Prior to the pandemic, telemedicine was investigated as a novel new way to provide post-surgical care to patients. Then the pandemic occurred, and it became a necessity rather than a novelty. When comparing the historical patients to those that received care during the pandemic, patients benefited from telemedicine care with no increase in 30-day healthcare encounters. Offering multiple options for post-discharge follow-up does not appear to have negative impact on healthcare utilization.
Funding
This research received funding from the American College of Surgeons Franklin H. Martin Faculty Research Fellowship.
Declaration of competing interest
The authors report no other financial relationships that would constitute a Conflict of Interest.
Acknowledgements
Sarah McCloy, Katie Collins, Debra Manning, Romeeka Perkins.
Appendix A. Supplementary data
The following is the Supplementary data to this article.
00102-2&id=ga1.jpg){kind=link}
00102-2&id=gr1.jpg){kind=link}
00102-2&id=gr2.jpg){kind=link}
00102-2&id=gr3.jpg){kind=link}