Abstract
Objectives
Patients with suspected COVID-19 remain at risk for clinical deterioration after discharge and may benefit from home oxygen saturation (SpO2) monitoring using portable pulse oximeter devices. Our study aims to evaluate patient engagement and compliance with a home SpO2 monitoring program.
Methods
This is a single center, prospective pilot study of patients being discharged from the ED or urgent care after evaluation of symptoms consistent with COVID-19. Subjects were given a portable pulse oximeter and instructed to obtain measurements at rest and with exertion twice daily for 14 days. Patients were contacted daily to collect recorded data. If attempts to contact the patient were unsuccessful for 3 consecutive days, patients were considered lost to follow up. The primary outcome of interest was patient engagement in the program which was defined as the percentage of patients that completed the 14-day study period, meaning they were not lost to follow up. Secondary outcomes included compliance with performing the SpO2 readings. Patient compliance was calculated as a percentage of completed readings out of the total expected readings.
Results
Fifty patients were enrolled - 2 withdrew and 1 was a screen failure. Overall, engagement in the program was 46.8% with no significant difference between those who tested positive for SARS-CoV-2 versus those who tested negative (48.2% vs 45%, p = 0.831). Median compliance overall was 42.9% (IQR 22.22-78.57). Median compliance for the positive group was 50.0% (IQR 20-85.71) and 42.86% (IQR 22.92-76.44) for the negative group (p= 0.838).
Conclusion
Our study demonstrated that there was acceptable engagement and compliance in a 14-day home SpO2 monitoring program. These results support the use of home pulse oximetry monitoring in a select group of mildly ill patients with suspected COVID-19.
Keywords: COVID-19, SARS-CoV-2, Pulse Oximetry, Hypoxemia, Home Monitoring
Public Interest summary.
The coronavirus disease 2019 (COVID-19) can present in a variety of ways. Patients typically experience viral symptoms including fevers, body aches, and dry cough although many can remain asymptomatic. When assessed in the emergency department, patients with no or minimal symptoms are usually discharged home and given recommendations regarding supportive care. Some patients with COVID-19 may develop severe decrease in oxygen levels before developing more serious signs and symptoms. Therefore, our emergency department acquired portable pulse oximeter devices to be given to patients being discharged and considered to be at risk of worsening health condition. Patients were instructed on the device use and asked to monitor their oxygen levels at home. Since home monitoring programs were new and becoming increasingly popular, the purpose of our study was to help inform whether patients are likely to follow instructions regarding home oxygen monitoring and use the pulse oximeter regularly.
Alt-text: Unlabelled box
Introduction
The novel coronavirus, SARS-CoV-2, which causes the clinical syndrome COVID-19 was declared a pandemic by the World Health Organization (WHO) on March 11, 2020.[1] Two days later, in the United States, a national emergency was declared.[2] Response to this pandemic has led to multiple innovations in the provision of care to patients such as the use of computer tablets for communication with patients,[3] deployment of outdoor tents to increase capacity for patient evaluation of care,[4] and the use of 3D printers to manufacture personal protective equipment (PPE).[5] There have been multiple surges in the number of patients requiring medical care for COVID-19. At times these surges can overwhelm the available resources and the capacity of hospitals. Space is a precious commodity during the current pandemic and therefore only patients with more moderate-to-severe disease are typically admitted to the hospital. Patients with no or minimal symptoms are usually discharged home and given recommendations regarding supportive care and monitoring of symptoms. Some patients with COVID-19 may develop dramatic reductions in oxygenation levels that may precede other signs and symptoms of clinical deterioration. This condition has been called “silent hypoxia.”[6]
Patient-performed evaluations have been used in a variety of conditions such as monitoring of blood glucose levels for patients with diabetes,[7] weight for patients with congestive heart failure,[8] and blood pressure for patients with hypertension.[9] The use of pulse oximetry devices has been described for conditions such as chronic obstructive pulmonary disease (COPD),[10] obstructive sleep apnea,[11] and congenital heart disease.[12] Pulse oximeter devices used for home monitoring are small, portable, and easy to use. These devices use near-infrared photospectroscopy and are based on the principle that hemoglobin has variable absorbance of light at different wavelengths based on the level of oxygenation.[13] Basic models report only the oxygen saturation (SpO2) through the measurement of light absorbance at two different wavelengths. More advanced models provide additional values such as heart rate and graphical depiction of the SpO2 waveform. Most models simply provide the data on a small digital screen; however, some are equipped to transmit the data to providers using a wi-fi internet connection.
Faced with concerns that large numbers of patients diagnosed with COVID-19 could be discharged home with the potential for decompensation without easily identified warning signs, our hospital acquired a large number of portable pulse oximeter devices to be distributed to patients considered to be at risk of decompensation due to COVID-19. Since this disease process was so novel, there was no data to support that these devices could identify patients sooner, lead to improved patient outcomes, and not result in over-utilization of resources due to false readings. Further, it was not known if patients provided with these devices would use them appropriately and as an adjunct to basic symptom monitoring.
Several studies have shown clinical benefits for those monitored closely in the outpatient setting with COVID-19. A study by Boniface showed a significant improvement in 30-day mortality for patients monitored at home with COVID-19 (Boniface, 2022). In another study, Shah et al utilized a 7-day home pulse oximetry program to follow up a cohort of 77 COVID-19 positive patients to identify need for hospitalization. Their subjects were instructed to return to the ED in case their home SpO2 dropped below 92% and the results showed that a home SpO2 < 92% was strongly associated with hospitalization (RR = 7; P < 0.0001).[14]
Although there was a clear need for outpatient monitoring of symptoms and vital signs in those diagnosed with COVID-19, it was not known whether our patient population would commit to the monitoring necessary for the outpatient program to be worthwhile. A study by Oliver et al showed that patients who participated in a home monitoring program for those with COVID-19 found the program “highly acceptable and easy to use” (Oliver, 2022). The question remains as to whether patients are willing to commit to an outpatient monitoring program and participate for an extended duration.[15]
The purpose of this study was to determine the level of engagement and compliance in a home pulse oximeter monitoring program for patients who were evaluated in the emergency department or urgent care center for symptoms concerning for COVID-19 but were considered appropriate for discharge at the time of evaluation. We hypothesize that the subjects in our study would comply with frequent monitoring of home vital sign measurements and symptoms assessments and thus, show an adequate level of participation in the study. The results of this study will help inform whether patients are likely to comply with instructions regarding home monitoring which will be useful in the development of large-scale home pulse oximetry monitoring programs.
Methods
Study design and setting
We conducted a prospective, non-randomized trial from May 31 until Oct 22, 2020 at the University of Maryland Medical Center Downtown Campus (UMMC) emergency department and the UMMC Urgent Care Center, both located in Baltimore, Maryland. During the study period, both sites were equipped to evaluate and test patients with symptoms concerning for COVID-19. The pulse oximeter devices used in this study were made available at no cost to the patient or study team through a hospital quality improvement initiative. There was no separate funding for this study. The protocol was approved by the local institutional review board (IRB) at the University of Maryland, Baltimore. All subjects enrolled in the study provided written informed consent.
Selection of Participants
The study population consisted of a convenience sample of adult patients, 18-years-old or older, seen in the ED or urgent care for symptoms consistent with SARS-CoV-2 infection but determined to be well enough to be discharged home after clinical assessment. To be eligible for inclusion, patients needed to have symptoms consistent with COVID-19, undergo testing for COVID-19, have a working phone number, and be willing to follow all study-related procedures. Pediatric patients (< 18 years of age), patients being admitted or placed in observation, patients who are pregnant or expressed concern for pregnancy, incarcerated patients, and patients who were otherwise noted as not suitable by the investigator or ED clinician were excluded from the study. Testing for SARS-CoV-2 at the enrollment sites was done using reverse transcriptase polymerase chain reaction (RT-PCR) of an oropharyngeal or nasopharyngeal swab. During the study period, various tests were used by the laboratory based on availability and most tests results were not available for up to 7 to 9 days after the specimen was collected. Due to the delay in receiving test results at the time of enrollment in the study, a positive SARS-CoV-2 test was not required for inclusion in the study.
Study sample size
This was a pilot study to determine preliminary data on the topic with a target enrollment of 50 patients. At the time of the design of this protocol, there had been no previously published studies assessing patient engagement in a home pulse oximetry monitoring program for COVID-19; as such, sample size calculation was not performed.
Intervention
Upon discharge from the ED, participants who signed the informed consent were given an Easy Grip Pulse Oximeter ™ (Medline, Northfield, IL, USA) at no cost and were instructed on its use. Patients were instructed to apply the pulse oximeter device on their finger and to record the results for pulse and oxygen saturation in their diary while at rest and with exertion (e.g., after a 6-minute walk). The day of discharge was considered the baseline visit of the study (Day 0). Each subject was given a diary form on which they would record data two times a day (morning and evening). The data collected included the pulse oximetry reading at rest and with exertion for a total of 14 days. Subjects were also asked to record the presence or absence of symptoms that may be associated with a SARS-CoV-2 infection. Patients were asked to record their daily temperatures on the diary form; however, thermometers were not provided. The choice of 14 days was selected to ensure enough follow-up time for the development of potential COVID-19 complications.
Measurements
Study team members called the subjects on a daily basis starting the day after enrollment to collect data recorded in the diary over the previous 24 hours or since last contact. If a call on a previous day was missed but the patient had recorded the values in the diary, the study team member collected the values but denoted the date and time that the values were actually obtained. The study team members also asked subjects about any contact with the healthcare system such as additional visits to a primary care provider or emergency department and if they were experiencing any issues with the devices or program overall. During each call, the study team member reminded the subject to seek immediate care at an ED if symptoms worsened. The purpose of the phone calls was to gather data and not to provide medical advice. If contact with a study subject was not successful on any day, an additional 3 attempts were made on the same day. If no contact was made with a study subject for 3 consecutive days, then the subject was considered lost to follow up and no further attempts to contact the patient were made. Subjects were allowed to withdraw from the study at any time.
Outcomes
Patient charts were reviewed to identify demographic data as well as past medical history. Information from the baseline visit was also collected retrospectively- this included symptoms, vital signs, laboratory findings, and imaging results. Obesity was defined as a body mass index (BMI) > 30 kg/m2. Categorical variables such as demographic characteristics of the study subjects are reported as counts and percentages. Continuous variables such as vital signs are reported as calculated medians (interquartile range).
The primary outcome of interest was the participants’ rate of engagement in the pulse oximetry program which was defined a priori as completion of the total duration of the study without loss to follow up or withdrawal from the study. Overall, engagement was calculated by dividing the number of participants who were considered engaged (i.e. not lost to follow up or withdrawn) over the total number of included participants. Secondary outcomes of interest included patient compliance using the pulse oximeter device as well as patient return to the hospital due to worsening clinical condition. Patient compliance over the duration of the program was calculated by dividing each patient's total reported pulse oximetry readings by the total expected readings. In order to have 100% compliance, a subject would be required to record the pulse oximetry reading every morning and evening for the duration of the 14-day enrollment period. For the sake of simplicity, and since many subjects were unable to complete the exertional readings due to fatigue, if a subject missed either a resting or an exertional reading, they were still considered compliant for that time point. Therefore, in order to have 100% compliance, it would be sufficient that a subject report either the resting or the exertional reading twice a day (morning and evening) for the duration of the 14-day period (28 readings in total).
Bias
To limit recall bias, subjects were instructed to not complete missed data entries based on memory. To account for an observer-expectancy bias, the study team members performing the phone call follow ups were blinded to the results of SARS-CoV-2 testing results until the analysis phase of the study. The same brand of pulse oximeters was used with all patients to minimize device variability in SpO2 values. Recently, concerns have been raised regarding the validity of pulse oximetry readings in patients with darker skin tones. No adjustments were made based on race or skin pigmentation.
Statistical analysis
The difference in the rate of engagement between COVID-19 positive and COVID-19 negative patients and the respective P-value was calculated using the Pearson Chi-squared test. Given the small sample size, the differences in continuous data endpoints such as compliance rates and ED laboratory values between both groups was performed using the Mann-Whitney U test. Missing data constituted less than 5% of the total data and were therefore excluded from the analysis without performing multiple imputation. Data analysis was performed using SPSS v.24. [IBM Corp., Armonk, N.Y., USA] and all p-values < 0.05 were deemed significant.
Results
From May 31st through October 8th, 2020, a total of 50 patients were enrolled during their index visit to the emergency department or urgent care center. Of the patients that were consented, only 47 patients were included in the primary analysis. One patient was considered a screen failure due to a change in the disposition during the index visit. Two patients withdrew from the study before providing data. All enrolled patients received a qualitative test for detection of SARS-CoV-2. Of the 47 subjects included in the analysis, 27 (57.4 %) tested positive (Figure 1 ).
Figure 1.
Flow diagram of patient enrollment and study outcomes.
Demographic and patient characteristics of our subject cohort are summarized in Table 1 . Our subjects had a median age of 39 (29-45) years, 59.6% were female, and 74.5% were Black (Table 1). Our cohort presented with a variety of comorbidities associated with a more severe course of COVID-19. The most common comorbidities observed in our study population were obesity (51.1%), diabetes mellitus (17.0%), asthma (17.0%), hypertension (14.9%), and dyslipidemia (6.4%). Comparing the COVID-19 positive and COVID-19 negative patients, most demographic and risk factors were well balanced, although the COVID-19 positive group was found to have a slight male predominance. A variety of data was collected during the patient's baseline visit including symptoms of COVID-19, vital signs, imaging findings, and laboratory data. The most commonly reported symptoms were cough (66%), shortness of breath (63.8%), and fever (46.8%). There were no significant differences between COVID-19 positive and COVID-19 negative groups in baseline symptoms, vital signs, or imaging findings with the exception of white blood cell count and serum calcium levels (Tables 2 and 3 ).
Table 1.
A comparison of patient characteristics between COVID-19 positive and COVID-19 negative groups
| Variable % (n) | All patients (n=47) | COVID-19 positive (n=27) | COVID-19 negative (n=20) |
|---|---|---|---|
| Median age (IQR) in years | 39 (29-54) | 40 (30-54) | 36.5 (28-53.8) |
| Gender | |||
| Male | 40.4 (19) | 51.9 (14) | 25 (5) |
| Female | 59.6 (28) | 48.1 (13) | 75 (15) |
| Race | |||
| African American | 74.5 (35) | 70.4 (19) | 80 (16) |
| Hispanic | 2.1 (1) | 3.7 (1) | 0 (0) |
| White | 17 (8) | 14.8 (4) | 20 (4) |
| Other | 6.4 (3) | 11.1 (3) | 0 (0) |
| Comorbidities | |||
| Obesity | 51.1 (24) | 48.14 (13) | 55 (11) |
| Diabetes Mellitus | 17 (8) | 14.8 (4) | 20 (4) |
| Dyslipidemia | 6.4 (3) | 7.4 (2) | 5 (1) |
| Hypertension | 14.9 (7) | 18.5 (5) | 10 (2) |
| Coronary artery disease | 2.12 (1) | 0 (0) | 5 (1) |
| Asthma | 17 (8) | 7.4 (2) | 30 (6) |
| Interstitial lung disease | 2.12 (1) | 0 (0) | 5 (1) |
Table 2.
Comparison of symptoms, vital signs, and imaging findings during index ED visit between COVID-19 positive and COVID-19 negative groups.
| Variable % (n) | All patients (n= 47) | COVID-19 positive (n=27) | COVID-19 negative (n= 20) | P-value |
|---|---|---|---|---|
| Symptoms on presentation | ||||
| Cough | 66 (31) | 59.3 (16) | 75 (15) | 0.355 |
| SOB | 63.8 (30) | 55.6 (15) | 75 (15) | 0.226 |
| Fever | 46.8 (22) | 55.6 (15) | 35 (7) | 0.238 |
| Chills | 23.4 (11) | 25.9 (7) | 20 (4) | 0.737 |
| Myalgias | 40.4 (19) | 44.4 (12) | 35 (7) | 0.561 |
| Headache | 23.4 (11) | 22.2 (6) | 25 (5) | 1.00 |
| Sore throat | 17 (8) | 14.8 (4) | 20 (4) | 0.707 |
| Chest pain | 17 (8) | 18.5 (5) | 15 (3) | 1.00 |
| Loss of smell | 12.8 (6) | 14.8 (4) | 10 (2) | 1.00 |
| Loss of taste | 12.8 (6) | 14.8 (4) | 10 (2) | 1.00 |
| Nausea/Vomiting | 19.1 (9) | 22.2 (6) | 15 (3) | 0.713 |
| Diarrhea | 23.4 (11) | 33.3 (9) | 10 (2) | 0.086 |
| Other symptoms* | 38.3 (18) | 40.7 (11) | 35 (7) | |
| Vital signs Median (IQR) | ||||
| SBP (mmHg) | 132 (118-151) | 132 (121-143) | 131.5 (117.25-158.25) | 0.614 |
| DBP (mmHg) | 84 (74-92) | 82 (72-90) | 84 (75-96.75) | 0.567 |
| HR (bpm) | 94 (80-106) | 96 (84-106) | 87 (74.75-105.75) | 0.256 |
| RR (/min) | 18 (16-20) | 18 (16-18) | 17.5 (16-20) | 0.732 |
| Temp C | 37.1 (36.8-37.9) | 37.3 (37-37.9) | 36.95 (36.8-37.38) | 0.117 |
| SpO2 (%) | 99 (97-99) | 98.5 (97-99.25) | 99 (98-99) | 0.593 |
| Imaging† | ||||
| Normal findings | 55.3 (26) | 51.9 (14) | 60 (12) | 0.560 |
| Imaging findings consistent with infectious process | 29.8 (14) | 37 (10) | 20 (4) | |
| Other findings | 8.5 (4) | 7.4 (2) | 10 (2) | |
| Imaging not done | 6.4 (3) | 3.7 (1) | 10 (2) |
Other symptoms: runny nose, fatigue.
Imaging: Imaging modalities include chest X ray and/or CT scan of the lungs. Infectious findings defined as hazy opacities, ground-glass opacities, patchy opacities, streaky opacities, consolidations as well as other findings that suggest a possible infection as per the radiology report. Other findings include cardiomegaly, atelectasis, congestion and other non-specific findings.
Table 3.
Comparison of serum laboratory values at index ED visit between COVID-19 positive and COVID-19 negative groups. Statistically significant values and bolded and labeled to be easily visualized.
| Lab variable Median (IQR) | Total (n=47) | COVID-19 positive (n=27) | COVID-19 negative (n = 20) | P-value |
|---|---|---|---|---|
| Complete Blood Counts | ||||
| Hemoglobin (g/dL) | 13.2 (12.1-14.2) | 13.25 (12.25-14.53) | 13.2 (11.5-13.9) | 0.416 |
| Hematocrit (%) | 40.8 (37.6-43.75) | 41.3 (38.38-44.18) | 39.2 (35.5-42.6) | 0.221 |
| WBC (K/mcL) | 5.5 (3.95-7.05) | 4.45 (3.48-6.15) | 6.9 (6.8-7.4) | 0.005* |
| Platelets (K/mcL) | 219 (186.5-270.5) | 217 (166.75-243.75) | 241 (203-315) | 0.143 |
| Serum Chemistries | ||||
| Sodium (mmol/L) | 138 (136-140) | 138 (136-140) | 138 (137-140) | 0.284 |
| Potassium (mmol/L) | 4 (3.6-4.3) | 4.05 (3.8-4.33) | 3.8 (3.6-4.3) | 0.345 |
| Chloride (mmol/L) | 102.5 (100.75-104.25) | 102 (100-105) | 103 (101-104) | 0.942 |
| Bicarbonate (mmol/L) | 26 (24-27) | 26 (24-27) | 27 (24-28) | 0.796 |
| BUN (mg/dL) | 9.5 (8-14.25) | 11 (8-15) | 9 (8-14) | 0.828 |
| Creatinine (mg/dL) | 0. 81 (0.69-0.98) | 0.91 (0.7-1.17) | 0.75 (0.69-0.86) | 0.378 |
| Glucose (mg/dL) | 99.5 (93-120) | 99 (92-120) | 103 (94-134) | 0.344 |
| Calcium (mg/dL) | 9.4 (9.05-9.75) | 9.2 (8.98-9.7) | 9.6 (9.3-10.1) | 0.038* |
| Phosphate (mg/dL) | 3.1 (2.95-3.45) | 3.1 (3.1-3.1) | 3.1 (2.9-3.6) | 0.807 |
| Total Protein (g/dL) | 7.85 (7.43-8.45) | 7.7 (7.3-8.3) | 8.1 (7.6-8.55) | 0.443 |
| Albumin (g/dL) | 4.3 (4.1-4.5) | 4.3 (4-4.4) | 4.45 (4.23-4.65) | 0.180 |
| AST (U/L) | 40 (26.5-51.5) | 43 (29-53.75) | 31 (26-38.5) | 0.095 |
| ALT (U/L) | 25.5 (16.75-41.25) | 29 (18.5-53.5) | 22 (15.5-30) | 0.122 |
| Alkaline Phosphatase (U/L) | 70.5 (53-84.75) | 70 (48.5-85) | 71 (56-91.5) | 0.608 |
| Bilirubin (mg/dL) | 0.5 (0.4-0.65) | 0.5 (0.4-0.7) | 0.6 (0.4-0.75) | 0.958 |
Overall engagement in the home pulse oximetry monitoring program was 46.8%. There was no significant difference in engagement between the COVID-19 positive and COVID-19 negative groups. Engagement was 48.2% and 45.0% for COVID-19 positive and COVID-19 negative groups respectively (p = 0.831). The secondary outcome of patient compliance was also found to be similar between groups. Overall median compliance was 42.9% (IQR 22.2-78.6) with compliance of 50.0% (IQR 20-85.71) and 42.9% (IQR 22.9-76.4) for COVID-19 positive and COVID-19 negative groups, respectively (p= 0.838). During the program period of 14 days, 25 patients were lost to follow up, of which 14 were COVID-19 positive and 11 were COVID-19 negative. The median number of days from patient notification of COVID-19 test results until loss to follow up was 7 (IQR 5-11) with similar distribution of 7 (IQR 5.5-13) and 7 (IQR 4-11) for COVID-19 positive and COVID-19 negative groups respectively.
In general, participant SpO2 readings remained within acceptable range for the duration of the study. Overall, median SpO2 was 98% for both resting and exertional readings. Median heart rate was 83 and 101 bpm for resting and exertional readings, respectively. There was no significant variation in SpO2 and HR between morning and evening readings.
Of note, 4 patients in the study cohort sought medical care during the study period; 3 were hospitalized while 1 presented to the ED twice without being admitted either time. All 4 patients seeking additional medical care were COVID-19 positive. Chief complaints on presentation included shortness of breath, diarrhea, and weakness. The hospital length of stay for these 3 patients ranged from 2 to 10 days. Prior to hospital admission, the average resting SpO2 reading of this cohort was 93% (range: 88%-98%) and the average resting heart rate was 86 beats per minute (range: 78-95).
Discussion
Our study was designed after several published articles discussed the notion of silent hypoxia and the potential role of home pulse oximetry monitoring for the earlier detection of clinical deterioration.[16] Depending on various socioeconomical factors, some patients may not be able to arrange close outpatient monitoring and would therefore benefit from home SpO2 monitoring. Recent studies reported on their success with phone calls without a predefined metric of engagement and compliance.[17] Other studies reported on the feasibility and patients’ subjective experience with a home SpO2 program which overall was positive.[15] To our knowledge, while there have been a few studies looking at the efficacy of a home pulse oximetry program for predicting unfavorable outcomes, our study is the first to investigate patient engagement and compliance in such a program.
We found that, overall, there was limited engagement (46.8%) and compliance (42.9%) with the pulse oximetry program. From our results, it did not seem that receiving a positive COVID-19 test result had a significant impact on the engagement and compliance rates. The reason behind this limited engagement is unclear however we speculate that in this ill population, using the pulse oximeter twice daily and writing down results may have been both time and effort intensive. Our method of investigating engagement was through daily phone calls, which requires consistency and commitment. Patients may not always be readily available to answer phone calls. Furthermore, patients may be reluctant to complete the exertional component of the program, particularly since many had experienced respiratory symptoms such as shortness of breath.
Although the available data on the use of an outpatient pulse oximeter has been promising, many experts have warned about intrinsic and extrinsic factors that may lead to inaccurate readings, particularly with the use of low cost or non-FDA approved pulse oximeter devices.[18] The device used in our study was an FDA-approved pulse oximeter that has a reported accuracy of +/- 2% for an SpO2 range of 70-100%. Overall, subjects reported that the device was simple, safe, and easy to use. There were no reported malfunctions. Participants also found that the study protocol was easy to understand although a few patients reported some difficulty primarily in the period immediately following enrollment. During day 1 of the program, 3 out of the 47 subjects (6%) missed their exertional readings due to difficulty understanding the program instructions. After clarification by a study team member, the same subjects had no difficulties with the program afterwards. Some patients reported that they were too fatigued to perform exertional SpO2 readings leading to missing values. However, as noted in the methods section, these patients would be considered to be compliant as long as the resting SpO2 was reported.
Our patient cohort was similar to the Shah study(14) in that they were clinically well enough to be discharged from the ED. Indeed, our sample was relatively young (Median age 39 years, IQR 29-54). There were no significant differences in the baseline ED visit data endpoints between the COVID-19 positive and negative groups. The median systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate (HR), respiratory rate (RR) and temperatures of the total sample of 47 patients were within the normal respective ranges. Similarly, the rest of the ED laboratory results of our patients were within the normal institutional laboratory ranges. This can be explained by our choice of a mildly ill sample population who were well enough to be discharged home. Despite being within normal limits, the median WBC count and serum calcium level were significantly higher in the COVID-19 negative group than the COVID-19 positive group (p = 0.005 and p = 0.038 respectively). While our study was not powered to look at laboratory differences between both groups, it is worth noting that these results are consistent with prior literature on leukopenia and hypocalcemia in the setting of COVID-19.[19], [20], [21]
Over the course of the study, the median SpO2, resting HR, and exertional HR of our participants were within normal limits (98%, 83 bpm, and 101 bpm respectively) which is also consistent with the nature of our mildly ill population. Unlike the study by Shah et al,[14] we did not set an SpO2 cutoff for instructing the patient to return to the ED. Our subjects were reminded on every phone call to report to the ED in the event of any worsening of their clinical condition. Of the four patients who sought medical care during our study, two of them returned to the ED after study team members identified a concerning trend of their SpO2 levels. One patient's resting SpO2 decreased from 95% on day 1 to 88% on day 5 while the second patient's resting and exertional SpO2 was approximately 86% during the first four days of the program. Despite the concerning SpO2 levels, both patients did not notice a change in the severity of their symptoms which included cough and shortness of breath. This prompted the study staff to inform the principal investigator of the study who in turn contacted the patients and urged them to seek medical attention. Both patients were then seen and assessed at an ED and were subsequently hospitalized. Although this data was not enough to infer statistically significant conclusions, we found our program to be helpful in identifying both of these unique cases who ended up requiring hospitalization.
Limitations
The total sample size included in the analysis of this study was 47 participants which may be a relatively small sample size to have an adequately powered analysis. Furthermore, the eligibility criteria were intentionally selected to ensure enrollment of only mildly ill patients with suspected COVID-19, thus the engagement and compliance rates found in the study may not be generalizable across a more severely ill COVID-19 population.
As part of the informed consent process, subjects were informed that the main goal of this study was to assess their engagement in the outpatient program and their compliance with pulse oximeter use. Additionally, daily phone calls may have served as a reminder for patients to be more compliant with taking their pulse oximeter measurements. Both factors may have introduced a Hawthorne effect where subjects had falsely higher engagement and compliance rates. An interesting finding was the significantly low rate of temperature recording. Only 2 patients out of the 47 (4%) consistently reported their temperature readings over the duration of the study. This may be due to the fact that thermometers were not provided to our patients upon enrollment. In addition, we did not perform a follow up assessment to determine the cause of loss to follow-up which may have potentially given us more insight about the patients’ perceived value of the program. Although patients have reported that the pulse oximeter devices were simple and easy to use, no objective metric of usability was used in our study.
Conclusions
This study found that patients with suspected COVID-19 had limited engagement and compliance rates of approximately 50% in a 14-day home pulse oximetry monitoring program. Overall, patients reported that the program design was easy to comprehend and that the pulse oximeters were simple, safe, and easy to use. Limited engagement and compliance may be explained by the time and effort - intensive design of our program. We recommend that future studies examine engagement with alternative methods of data collection, including the use of cell phone applications and real time remote SpO2 monitoring.
Although not initially among the objectives or scope of our study, the program identified two unique patients in whom pulse oximetry monitoring helped in the early detection of hypoxia without concurrent worsening in symptoms, which then led to their hospital admission. We believe that the results of our pilot study should encourage the judicious use of home pulse oximetry monitoring by frontline providers in a select group of mildly ill patients with suspected COVID-19. If implemented on a wide scale, this may not only help detect early hypoxia and improve patient-centered outcomes, but also decrease unnecessary hospital utilization, crowding of emergency departments, and use of personal protective equipment. Further study with different program designs is needed to determine ways in which engagement and compliance can be improved.
Funding Sources/Disclosures
None
Consent
Informed consent process was conducted throughout the enrolment process. Due to COVID-19, the original physical copy of the informed consent remained with the patient and a digital copy was uploaded into their electronic charts.
Ethical Approval
This study has been approved by the IRB at the University of Maryland, Baltimore.
Meetings
The results of this study have been presented in oral abstract form at the ACEP Research Forum on October 27, 2021.
CRediT authorship contribution statement
R. Gentry Wilkerson: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration. Youssef Annous: Conceptualization, Methodology, Investigation, Writing – original draft, Writing – review & editing. Eli Farhy: Conceptualization, Methodology, Investigation, Writing – original draft, Writing – review & editing. Jonathan Hurst: Conceptualization, Methodology, Investigation, Writing – original draft, Writing – review & editing. Angela D. Smedley: Conceptualization, Methodology, Supervision, Writing – review & editing.
Declaration of Competing Interest
None declared
References
- 1.World Health Organization . World Health Organization [Internet]; 2020. WHO Director-General's opening remarks at the media briefing on COVID-19 - 11 March 2020.https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid Mar 11 [cited 2023 Jun 21]. Available from: -19—11-march-2020. [Google Scholar]
- 2.Federal Register Declaring a national emergency concerning the novel coronavirus disease (COVID-19) outbreak. Federal Register [Internet] 2020 https://www.federalregister.gov/documents/2020/03/18/2020-05794/declaring-a-national-emergency-concerning-the-novel-coronavirus-disease-covid-19-outbreak Mar 18 [cited 2023 Jun 21]. Available from: [Google Scholar]
- 3.Fang J, Liu YT, Lee EY, Yadav K. Telehealth Solutions for In-hospital Communication with Patients Under Isolation During COVID-19. West J Emerg Med. 2020;21(4):801–806. doi: 10.5811/westjem.2020.5.48165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Peterson KHJ, Muckey EJC. Deployment and Operation of Outdoor Treatment Tents During the COVID-19 Pandemic. Disaster Med Public Health Prep. 2020:1–4. doi: 10.1017/dmp.2020.355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wesemann C, Pieralli S, Fretwurst T, Nold J, Nelson K, Schmelzeisen R, et al. 3-D Printed Protective Equipment during COVID-19 Pandemic. Materials (Basel) 2020;13(8) doi: 10.3390/ma13081997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Wilkerson RG, Adler JD, Shah NG, Brown R. Silent hypoxia: A harbinger of clinical deterioration in patients with COVID-19. Am J Emerg Med. 2020;38(10):2243. doi: 10.1016/j.ajem.2020.05.044. e5- e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Olczuk D, Priefer R. A history of continuous glucose monitors (CGMs) in self-monitoring of diabetes mellitus. Diabetes Metab Syndr. 2018;12(2):181–187. doi: 10.1016/j.dsx.2017.09.005. [DOI] [PubMed] [Google Scholar]
- 8.Arulnathan A, Vaaheesan S, Denecke K. A Mobile Application for Self-Monitoring for Patients with Heart Failure. Stud Health Technol Inform. 2019;259:113–116. [PubMed] [Google Scholar]
- 9.Whelton PK, Carey RM, Aronow WS, Casey DE, Jr., Collins KJ, Dennison Himmelfarb C, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension. 2018;71(6):1269–1324. doi: 10.1161/HYP.0000000000000066. [DOI] [PubMed] [Google Scholar]
- 10.Minami S, Yamamoto S, Ogata Y, Nakatani T, Takeuchi Y, Hamaguchi M, et al. Ambulatory pulse oximetry monitoring in Japanese COPD outpatients not receiving oxygen therapy. Multidisciplinary Respiratory Medicine. 2014;9(1):24. doi: 10.1186/2049-6958-9-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mashaqi S, Staebler D, Mehra R. Combined nocturnal pulse oximetry and questionnaire-based obstructive sleep apnea screening - A cohort study. Sleep Med. 2020;72:157–163. doi: 10.1016/j.sleep.2020.03.027. [DOI] [PubMed] [Google Scholar]
- 12.Martin GR, Ewer AK, Gaviglio A, Hom LA, Saarinen A, Sontag M, et al. Updated Strategies for Pulse Oximetry Screening for Critical Congenital Heart Disease. Pediatrics. 2020;146(1) doi: 10.1542/peds.2019-1650. [DOI] [PubMed] [Google Scholar]
- 13.Sinha IP, Mayell SJ, Halfhide C. Pulse oximetry in children. Arch Dis Child Educ Pract Ed. 2014;99(3):117–118. doi: 10.1136/archdischild-2013-305526. [DOI] [PubMed] [Google Scholar]
- 14.Shah S, Majmudar K, Stein A, Gupta N, Suppes S, Karamanis M, et al. Novel Use of Home Pulse Oximetry Monitoring in COVID-19 Patients Discharged From the Emergency Department Identifies Need for Hospitalization. Acad Emerg Med. 2020;27(8):681–692. doi: 10.1111/acem.14053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Oliver J, Dutch M, Rojek A, Putland M, Knott JC. Remote COVID-19 patient monitoring system: a qualitative evaluation. BMJ Open. 2022;12(5) doi: 10.1136/bmjopen-2021-054601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Jouffroy R, Jost D, Prunet B. Prehospital pulse oximetry: a red flag for early detection of silent hypoxemia in COVID-19 patients. Crit Care. 2020;24(1):313. doi: 10.1186/s13054-020-03036-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gootenberg DB, Kurtzman N, O'Mara T, Ge JY, Chiu D, Shapiro NI, et al. Developing a pulse oximetry home monitoring protocol for patients suspected with COVID-19 after emergency department discharge. BMJ Health Care Inform. 2021;28(1) doi: 10.1136/bmjhci-2021-100330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Quaresima V, Ferrari M. COVID-19: efficacy of prehospital pulse oximetry for early detection of silent hypoxemia. Crit Care. 2020;24(1):501. doi: 10.1186/s13054-020-03185-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Di Filippo L, Formenti AM, Rovere-Querini P, Carlucci M, Conte C, Ciceri F, et al. Hypocalcemia is highly prevalent and predicts hospitalization in patients with COVID-19. Endocrine. 2020;68(3):475–478. doi: 10.1007/s12020-020-02383-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Liu J, Han P, Wu J, Gong J, Tian D. Prevalence and predictive value of hypocalcemia in severe COVID-19 patients. J Infect Public Health. 2020;13(9):1224–1228. doi: 10.1016/j.jiph.2020.05.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mina A, van Besien K, Platanias LC. Hematological manifestations of COVID-19. Leuk Lymphoma. 2020;61(12):2790–2798. doi: 10.1080/10428194.2020.1788017. [DOI] [PubMed] [Google Scholar]