International health organizations and officials are bracing for a pandemic. Although the 2003 severe acute respiratory syndrome (SARS) outbreak in Toronto did not reach such a level, it created a unique opportunity to identify the optimal use of the Internet to promote communication with the public and to preserve health services during an epidemic.
The aim of the study was to explore patients’ attitudes regarding the health services that might be provided through the Internet to supplement those traditionally available in the event of a future mass emergency situation.
We conducted “mask-to-mask” surveys of patients at three major teaching hospitals in Toronto during the second outbreak of SARS. Patients were surveyed at the hospital entrances and selected clinics. Descriptive statistics and logistic regression models were used for the analysis.
In total, 1019 of 1130 patients responded to the survey (90% overall response rate). With respect to Internet use, 70% (711/1019) used the Internet by themselves and 57% (578/1019) with the help of a friend or family member. Of the Internet users, 68% (485/711) had already searched the World Wide Web for health information, and 75% (533/711) were interested in communicating with health professionals using the Internet as part of their ongoing care. Internet users expressed interest in using the Web for the following reasons: to learn about their health condition through patient education materials (84%), to obtain information about the status of their clinic appointments (83%), to send feedback to the hospital about how to improve its services (77%), to access screening tools to help determine if they were potentially affected by the infectious agent responsible for the outbreak (77%), to renew prescriptions (75%), to consult with their health professional about nonurgent matters (75%), and to access laboratory test results (75%). Regression results showed that younger age, higher education, and English as a first language were predictors of patients’ interest in using Internet services in the event of an epidemic.
Most patients are willing and able to use the Internet as a means to maintain communication with the hospital during an outbreak of an infectious disease such as SARS. Hospitals should explore new ways to interact with the public, to provide relevant health information, and to ensure continuity of care when they are forced to restrict their services.
International public health organizations and officials around the world are bracing for a pandemic [
The 2003 severe acute respiratory syndrome (SARS) outbreak caused major disruption of hospital services in affected countries [
Although effective, telephone-based communication proved to be a resource-intensive solution that may not be sustainable in the event of a more widespread epidemic or external disaster. Considering that most Canadians have access to the Internet [
After obtaining permission from the UHN’s Command Centre and approval from the institutional Research Ethics Board, we conducted a cross-sectional survey at the single access points of each of the hospitals comprising UHN (Toronto General Hospital, Toronto Western Hospital, and Princess Margaret Hospital) and six of the ambulatory clinics that remained open during July 2003.
The survey was based on the core questions of the UHN e1000 study, a cross-sectional survey exploring the patterns of Internet use among patients and providers associated with the UHN. The e1000 survey [
Screening during SARS at the University Health Network
To alleviate possible anxiety during the crisis, only 9 questions with categorical responses were asked, in addition to those capturing demographic information (
The initial plan to survey all patients entering the hospital after being screened for SARS (approximately 1000 patients daily per hospital) proved to be unrealistic due to the cumbersome process of registration and screening that created additional challenges for those people approaching patients. For three days, we surveyed patients after they passed the entrance screening points, and we concluded that the clinics’ waiting rooms would be a more appropriate venue for recruiting. Only adult ambulatory patients were approached. Ten trained multilingual interviewers conducted the surveys in English after obtaining verbal consent. The interviewers and patients were required to wear full protective gear, which meant “mask-to-mask” rather than face-to-face communication (
With no face recognition and to avoid re-approaching the same patients, interviewers used colored stickers on the masks of patients to identify those already invited to participate in the survey.
Contact information provided voluntarily by patients was stored in a secure database on the Centre for Global eHealth Innovation servers. Hard copies of the surveys were stored in locked cabinets and were accessible to researchers for analysis only at the Centre.
Descriptive statistics were gathered for each of the answers. Multiple logistic regression analysis was used to assess the effect of the sociodemographic variables on patients’ views of Internet usage for specific services in the event of a mass emergency. We used stepwise, forward, and backward methods with all variables to specify which ones stayed in the model. A cutoff
To ensure better reporting, we used the relevant items of the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) [
Protective gear for staff and interviewers during the SARS outbreak at UHN
A total of 1130 patients were approached, and 1019 chose to complete the survey, giving an overall participation rate of 90%. Refusal rate was initially higher at the entrances of the hospitals during the first three days of the survey (response rate 78%, 309/396 respondents approached) than later at the clinics that remained open (response rate 97%, 710/734 respondents approached). The most frequent reasons for declining were being late for a clinic appointment (50%), frustration and exhaustion due the long lineup to enter the hospital (20%), lack of interest (20%), and inability to speak English (10%).
We found that 91% of patients were aware of the Internet (926/1019) and that 70% used the Internet (711/1019) by themselves and 57% (578/1019) with the help of a friend or family member. Of the Internet users, 68% (485/711) had already searched the World Wide Web for health information, and 75% (533/711) were interested in using the Internet to communicate with health professionals as part of their ongoing care.
Table 1 shows the demographic characteristics of all respondents, those with Internet access, and by survey location (entry point or clinic). Overall, there was a balanced gender representation of Internet users in our sample, with 44% (313/711) women and 45% (320/711) men. The majority of Internet users were in the 21 to 40 (32%, 224/711) and 41 to 60 (40%, 281/711) age categories. Almost half (42%, 300/711) of the Internet users had college or undergraduate education, 3 out of every 4 Internet users (558/711) spoke English as their first language, and 64% of them (453/711) were born in Canada. These demographic proportions are comparable for all users and for users at the entry doors or clinics.
There were no statistically significant differences between Internet users surveyed at entry doors and clinics with respect to age (
Demographic characteristics
Male |
44% (451/1019) |
45% (320/711) |
38% (74/197) |
48% (246/514) |
< 21 |
2% (21/1019) |
3% (21/711) |
1% (3/197) |
3% (18/514) |
Elementary |
5% (49/1019) |
1% (9/711) |
1% (3/197) |
1% (6/514) |
English |
72% (734/1019) |
78% (558/711) |
78% (154/197) |
79% (404/514) |
Canada |
57% (583/1019) |
64% (453/711) |
62% (122/197) |
64% (331/514) |
* Percentages may not add to 100% for each variable due to missing responses.
In the event of a future outbreak, Internet users expressed interest in accessing the Internet to learn about their health condition through patient education materials (84%, 594/711), to obtain information about the status of their clinic appointment (83%, 590/711), to send feedback to the hospital about how to improve its services (77%, 549/711), to access screening tools to help determine if they were affected by the infectious agent responsible for the outbreak (77%, 544/711), to renew prescriptions (75%, 535/711), to consult with their health professional about nonurgent matters (75%, 536/711), and to obtain laboratory results (75%, 534/711). Respondents had the opportunity to suggest other uses for the Internet, and 10% (70/711) chose to do so. Their most frequent suggestion was the ability to communicate with family members, as visits were restricted. Others wanted to use the Internet to access their electronic health record, participate in virtual support groups, replace certain follow-up visits with online consultations, and find information on drug compatibility or clinical trials.
Statistically significant demographic predictors for interest in specific Internet services among Internet users are shown in
Logistic regression of demographic factors (independent variables: rows) predicting interest in specific Internet services (dependent variable: columns) among Internet users (n = 711)
|
|||||||||||||
|
|
|
|
|
|
|
|
||||||
|
|||||||||||||
< 40 | 1.98 |
2.37 |
2.12 |
1.98 |
1.96 |
2.27 |
2.51 |
2.05 |
|||||
41-60 | 1.38 |
1.93 |
1.76 |
1.84 |
2.23 |
2.18 |
1.61 |
2.21 |
|||||
> 60 (RC) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |||||
|
|||||||||||||
High school or less (RC) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |||||
College/ |
1.79 |
2.34 |
2.19 |
2.04 |
1.67 |
1.55 |
1.72 |
NS | |||||
|
|||||||||||||
No (RC) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |||||
Yes | 1.87 |
2.45 |
2.13 |
2.73 |
1.36 |
1.38 |
1.96 |
NS | |||||
|
|||||||||||||
Female (RC) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |||||
Male | 0.75 |
NS | NS | NS | NS | NS | 1.33 |
NS | |||||
|
|||||||||||||
No (RC) | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | |||||
Yes | NS | NS | 0.58 |
0.54 |
NS | NS | NS | NS |
RC = reference category
NS = nonsignificant factors (
Internet users (
The detailed results for the populations at the entry doors and at the clinics are provided in
At the clinics (Table 4 in
Overall, younger age, higher education, and English as a first language were predictors of interest in using Internet services in the event of a pandemic, with a few exceptions.
Four people in Toronto died of SARS, while hundreds were infected around the world. However, the SARS outbreak pales in comparison to a full-blown pandemic. For instance, the bubonic plague killed more than 130 million people, while the Spanish flu pandemic of 1918 killed more than 30 million. In Philadelphia, the 1793 yellow fever outbreak took the lives of more than 4000 people. Today, research suggests that the world is due for a pandemic [
The 2003 SARS outbreak challenged the way in which health organizations deal with public health crises. Although the classic outbreak control measures (infection control, contact tracing, quarantine, etc) were used in order to overcome new obstacles, such as high volume of air travel, increased media attention, and generalized panic, alternative methods of communication and collaboration to overcome them were required.
Similar to what happened during the anthrax scare [
Our results suggest that most patients are willing and able to use the Internet as a means to preserve and complement hospital information and communication services during an outbreak of an infectious disease such as SARS.
The results of this study are consistent with previously conducted surveys at UHN during non-SARS times. These results are related only to the proportion of patients using the Internet for general and health purposes. Earlier iterations indicate that 60% of respondents have used the Internet for general purposes and 69% for retrieving health information [
We conducted a systematic review of the literature looking for surveys of patients in relation to the type of services desired but did not find similar enough studies to justify a comparison with our SARS survey (data not shown).
There are many other potential uses for the Internet as a means of communication if hospitals and clinics were disabled by a new outbreak. Members of the public with Internet access who are quarantined may use it to get answers to nonurgent questions related to the infectious disease or to receive reassurance that they are managing their health properly [
Harnessing the power of the Internet in the event of a new outbreak, and particularly during a pandemic, will require changes at the hospital level that need to be gradually introduced during “new normal” times. At the very least, as part of the patients’ registration process, hospitals should collect data regarding patients’ choice for communication method (telephone, email, or both) in the event of an outbreak.
The SARS crisis underscored many opportunities for the use of Internet-mediated communication to extend the continuum of care outside of hospital walls, even under normal circumstances. Embracing the Internet as an integral part of clinical care, however, will require changes in legislation, funding structures, and flexible work patterns to enable health professionals to use it [
The findings of our survey highlight the need for timely, relevant, valid, feasible, and substantiated options to maintain communication lines with the public during crises that disable hospitals. We are aware that Internet access is not yet universal, but it certainly could be very valuable for the large subset of the population that uses it [
With the current increased risk of pandemics and bioterrorist attacks, it is essential to put in place the mechanisms necessary to use the Internet effectively and efficiently in order to reduce the impact of these crises on the health system and the public at large.
The special circumstances under which these surveys were conducted presented several design, execution, and data analysis limitations.
The sample chosen was one of convenience. Due to constraints inherent to the emergency, it was difficult to ensure that all individuals attending the hospital at any given date had the same probability of being selected. The difficulty of obtaining an up-to-date list of patients visiting the hospital limited our ability to establish an accurate denominator. Therefore, the likelihood of the sample being representative of the population attending the hospital during the second outbreak of SARS is unknown. To reduce the evident sampling bias, a number of random sampling techniques could have been used. Assuming that the population at selected clinics was captive and had patients with similar health conditions, cluster sampling may have been indicated.
At the hospital entry points, some interviewers noted that respondents were rushing and may have fallen into a rut (a “response set”), continuing to give the same response unthinkingly. This was particularly evident for question 6, which may have elicited repeated identical responses (ie, yes). In addition, the number of refusals at the single point entrances was higher than previously obtained at UHN [
Comparisons of participants from the three participating hospitals and an assessment of users versus non-Internet users were outside the scope of this study. Furthermore, the patient’s health condition was not collected. For future studies, these elements will be considered.
We are in indebted to the UHN Command Centres at each of the hospitals, the interviewers, and the staff who made this study possible. Funds for this study were provided by the Centre for Global eHealth Innovation (DL, AJ, HB, CR), the Canada Research Chair in eHealth Innovation (AJ and DL), and the Rose Family Chair in Supportive Care (AJ and CR).
CR and DL designed and carried out the study, led the analysis and interpretation of data, and drafted and revised the manuscript. HB provided statistical expertise and contributed to the analysis. MA and TC contributed to the concept of the study and revised the manuscript. AJ contributed to the concept and design of the study, interpreted the data, and revised the manuscript. All authors read and approved the final manuscript.
None declared.
Study survey.
Additional tables (Tables 3 and 4) showing the results of the populations at the entry doors and clinics.