Main findings
Our study shows that patient satisfaction for the survey population with the present TM service was fairly high with 78% of the respondents answering that they would use PVO for similar symptoms in the future and an overall satisfaction rate of 79 (0–100 scale). Factors significantly associated with high satisfaction were lower age and being assessed by a doctor instead of a nurse.
Most of the respondents (82%) would have visited another health care provider if they had not gotten the PVO appointment the same day. About three-quarters (74%) of the respondents did not intend to seek more health care for the same cause during the coming week.
The high satisfaction with TM is in-line with physical care visits in previous research where around 80% would respond yes to the question if they would recommend their PHC to someone in a similar situation [
12]. The oldest age group was less likely to use PVO again and they were also least satisfied with their visit. It has earlier been shown that personal interaction and continuity are more important for older patients than time or travel convenience and that they are not as familiar with describing symptoms in digital format as younger patients [
13]. This could be the reason for a lower satisfaction level for the oldest age group.
Our results show that the survey respondents were more likely to use PVO again if assessed by a doctor compared to a nurse. Earlier research has described patients using TM as being confident in knowing that TM was the right level of care for them [
4]. Therefore, one possible explanation is that the nurse’s assessment of patients in less need of health care is not in-line with the patient’s own perception. If users’ expectation of contact with a doctor is not met, this could affect the likelihood to use the service again [
14]. The finding that patient satisfaction was lower for patients assessed by a nurse is opposite to the results from a Cochrane systematic review of 18 randomised trials in 2018. This showed that patient satisfaction in primary care was slightly higher for nurse-led care for patients with all types of health problems, excluding mental health problems, regarding both first contact and ongoing care [
15]. This discrepancy between our findings that patient satisfaction when assessed by a nurse was lower in TM further emphasises the need for more research focused on TM.
Besides the statistically significant findings, we found some trends in certain areas, but where further research is needed. It seemed like the respondents under the age of 20 were more likely to use PVO again than the rest of the population. Earlier research has shown that time and accessibility is important for parents with children in need of health care. By using TM, they can cut travel time and increase their flexibility [
4]. We also noted differences between the symptom groups where the overall satisfaction level was the highest for gastrointestinal symptoms, which further suggests that certain subcategories are more suitable for TM.
Strengths and limitations
One strength of the study is that age and sex distribution seem to be in-line with the overall demographic characteristics of patients visiting PVO during the timespan, indicating that the survey cohort is representative of the whole patient group.
Another strength is that the survey was accessible to patients during both summer and winter. In that way, we could explore patient satisfaction over a larger time span and not only during typically seasonal health issues.
A major limitation of this study is the low response rate. Web-based surveys are known to have a significantly lower response rate than paper-based [
16]. A Danish randomised study showed a risk difference of 55.3% in response rate for the internet group compared to the paper-based for a survey distributed to patients with breast cancer. The response rate for the Danish web-based survey was 17.9%, which is a lot higher than ours but still considered to be low [
17]. A reason for this may be that a cancer patient is more prone to answer questions about their treatment than a patient with a minor problem in primary care, and hence our response rate is even lower. Although, even web-based surveys to cancer patients may have as low response rate as 5.5% [
18].
A way to increase the response rate would have been through reminders [
19]. The design of this study did not enable this with the survey in a pop-up window not connected with patient data.
However, despite the low response rate, the evidence shows comparable or even better reliability with web-based surveys compared with paper-based surveys [
20]. It is also possible that patients want to justify their digital visit by saying they would have sought another health care provider anyway [
21]. The survey was intended to be short and easy to complete, although further and more detailed questions could have provided more knowledge to understand the patients’ satisfaction.
We do not believe that it is possible to ignore the nonresponse bias in our study. The mechanisms that result in nonresponse can only be speculated. As patients who are extremely positive/negative are more prone to answer surveys [
22], one can believe that the respondents to the survey want to express something either positive or negative about their experience. The nonrespondents may have been more neutral in their responses. We are not able to measure the relationship between the opinions of the different groups, since the required data is unavailable. Our speculations about the nonresponders being more neutral in their opinions about the PVO service make us believe that the costly pursuit of a high response rate may had offered little or no reduction in nonresponse bias. There are statistical procedures to correct for the sampling bias introduced by nonresponse, but as stated by Burkell et al. [
23], unfortunately these statistics work best when they are needed least: at low levels of nonresponse, and hence would not be possible to use for this material.
However, we do have almost 700 responses and we believe that these opinions are important to show, as we are dealing with a relatively new phenomenon of telemedicine. For further development of the service, it is important to acknowledge the respondents’ opinions.
A further limitation of this study is that it only covers one region of Sweden. The care delivered from PVO does though serve patients from multiple PHCCs and is provided to patients throughout the whole region, covering a rather large geographical area with a diverse population regarding sociodemographic and rural/urban determinants. However, in future studies, it would be beneficial to involve digital providers in other regions.
Comparison with other studies
Several previous studies have focused on the impact of TM on accessibility, as well as staff and patient satisfaction. A systematic review from 2016 concluded that patients were generally satisfied with TM and the satisfaction rate was consistently above 80% and comparable with in-person visits [
24].
The satisfaction rate of the digital visits of PVO was in-line with the results from the Swedish National Patient survey 2021 where 78.5% (
N = 10,309) of respondents in Skåne would recommend their physical PHCC to someone in a similar health situation [
12]. It is also consistent with previous research showing that satisfaction rates for TM have been comparable with in-person visits [
24] and that users of TM were generally quite satisfied [
4,
24].
Most of the patients were women, which is in-line with previous studies that have shown that women use TM more often than men [
1,
4,
25]. Our results show that 82% would have used other health care providers if they would not have come in contact with PVO. These results are in-line with a study from 2018 showing 78% [
26] would have used other health care providers if they had not used TM for the same cause. This indicates that TM may unburden physical health care.
A systematic review of 18 randomised trials from 2018 showed that patient satisfaction is probably slightly higher in physical nurse-led primary care than in doctor-led primary care [
15]; this contradicts our main finding that patients were more satisfied when assessed by a doctor in TM care.
A study from the US in 2016 showed that the primary motivation for using telehealth was shorter waiting times [
25]. One-third of the patients in this study - and particularly those with no health insurance - preferred a TM visit to a traditional visit [
25]. Accessibility to primary health care could also have been important for our population but was not explored further in the survey, but since Swedish health care is largely governmentally funded, affordability would not have been a large issue for this survey population.