Effectiveness of a community paramedic-led health assessment and education initiative in a seniors’ residence building: the Community Health Assessment Program through Emergency Medical Services (CHAP-EMS)

A prospective pre-post approach was used for this intervention study. The CHAP-EMS program was implemented from January 1, 2013 to December 31, 2013. The program was piloted between November and December 2012 to ensure fidelity of the program before the full evaluation was conducted.

The study took place in a seniors’ housing building in Hamilton (a city in Ontario, Canada, population 520,000), [11] which was comprised of 260 apartments units, 84.5% of which were occupied by seniors aged 65 or older. Most occupants were assessed as being of low income and, accordingly, were receiving rent subsidies from the City of Hamilton Housing department. The building was identified by Hamilton Paramedic Services as having an unusually high EMS call rate and was considered an ideal location for the placement of the community health assessment program. Eligible study participants were all building residents over 65 years old. The program was delivered in a community common room area at a small informal station.

The CHAP-EMS program was delivered by two trained paramedics for eight hours a week (four hours per session, both sessions were on the same day). A more detailed description of our intervention is published elsewhere. [10] The paramedics were trained by a public health nurse and family doctor to take blood pressure, conduct health assessments, educate residents, and utilize the CHAP-EMS database. Building residents were invited to participate in the CHAP-EMS program through posters/flyers and announcements in tenant meetings delivered by members of the research team and paramedics. After providing written consent, participants were guided through a 15–20 min defined risk assessment by a trained paramedic. Risk factors assessed were those related to cardiovascular and diabetes risk (BP, diabetes-risk status, lifestyle factors), and potential for falls. Blood pressure was taken using the automated WatchBP Office (which takes 3 measurements on the same arm, on seated participants, and averages the results) following the Hypertension Canada recommended procedure [12]. All information was entered in the CHAP-EMS database which summarized the risk factor information based on pre-specified algorithms. Based on these, the paramedic provided education and developed an individualized action plan directing participants to use available community resources to assist them in addressing their risk factors. All participants received a copy of their risk profile, relevant health education, and information on local resources. They were advised to return to CHAP-EMS sessions regularly for BP monitoring and follow-up, and to further address risk factors in an ongoing fashion. Each participant’s information was faxed to his/her family physician once a month. Participants were also given cards containing a record of all BP readings to use for self-monitoring and sharing with their family physicians. Participants who had moderate or high risk of developing diabetes were advised to return for a fasting capillary blood glucose (CBG) test at the next visit. Participants were advised to see their physician if they had high CBG test results. A standard protocol was followed for referral or emergency action when participants had alarming findings (very high or low BP or CBG).

BP was collected using the Watch-BP Office (a Canadian Hypertension Society validated machine). Diabetes risk was generated using the Canadian Diabetes Risk (CANRISK) questionnaire [13], which asks 12 questions and will determine a risk category of either low, moderate or high for developing diabetes in the next 10 years. Cardiovascular risk factors were collected by asking about lifestyle details, such as dietary habits (including salt intake, fast food and fat intake). These questions were obtained from the established CHAP Program of research [14], and standard Canadian Community Health Survey. [15] Risk of falls was assessed using the Timed Up and Go (TUG) test [16]. Demographic information, BP values, CANRISK scores, and data about modifiable risk factors as recorded in the CHAP-EMS database were collated for each individual, from each session. EMS emergency (911) call rates specific to the intervention building were obtained from the routinely collected statistics held by the Hamilton Paramedic Service for the study year (2013) and the two preceding years (2011, 2012) to allow trends to be examined. This information was collected by the 911 call dispatch centre by postal code, as well as separately on forms that each ambulance team completed after a call. These two sources of data were aggregated and reconciled.

The power was computed post-hoc since the study was initially conducted to determine the feasibility of building resident participation in the program. The intention was to detect a change of 5 mmHg in both the systolic and diastolic BP (SD = 10 mmHg) because this has clinically significant implications [17]. With the 79 residents who participated in the study, there was a power of 0.88 to detect the desired change in BP.

All seniors who were 65 years and older and visited the CHAP-EMS program at least once were included in the analysis for the BP and CANRISK changes. Although the target of the CHAP-EMS program was older adults, any building resident who attended the program was assessed and received feedback, but data for those younger than 65 years were not analyzed. The change in the EMS emergency (911) call rates was based on all calls coming from the subsidized building whether or not the callers participated in the CHAP-EMS study, so providing the program to younger residents could have had an effect on this outcome. Data regarding call numbers was collected prospectively during the implementation period, and had been retrospectively collected for the two year period prior to implementation; duplicate calls were removed and individual visits only were accounted for in the statistical analysis.

Demographic characteristics and modifiable risk factors of the attendees were analyzed using descriptive statistics. Hierarchical linear modelling of BP was conducted to assess for the changes in systolic and diastolic BP over time within the same individual. We also analyzed the number of visits for normalisation of BP according to the 2015 Canadian Hypertension Education Program Guidelines, aiming for the goal of 140/90 [18]. Baseline descriptive analysis of the CANRISK scores was performed, as well as a comparison with the descriptive analysis of CANRISK scores obtained 6 months after the initial visit. Preliminary post-hoc cost analysis, including a sensitivity analysis, was performed using referenced costs from the literature for ambulance transport [19, 20], EMS time and hospital visits [21‐23], associating these with reductions in EMS calls, to assist in the interpretation of the results.

Table 1 summarizes the participant characteristics and results of the risk factor assessments. The mean age of the participants was 72.2 years. Most participants were female (68.1%) and 90.3% had a family doctor. The number of medications taken by the participants daily ranged from 0 to 12 medications (Mean 6.7, SD 4.2).

In 2011, prior to any program implementation, 114 emergency (9-1-1) calls were received from the study site. In 2012, a modified version of the program looking only at BP had run for a few months in the latter portion of the year, as the program was developed and refined; the number of emergency calls received was 104. In 2013, the CHAP-EMS program was fully implemented for the duration of the entire year and the number of emergency calls dropped to 84. Compared to the average number of emergency calls in the building in the previous 2 years (2011 and 2012), the number of emergency calls decreased by 7.1% (CI: 3.1–12.5%) at the 6 month mark, and 25% (CI: 17.3–34.0%) after a full year of CHAP-EMS programming (2013).

At baseline, the most prevalent risk factors displayed by the participants were high waist circumference (63.9%) and elevated body mass index (61.1%). These risk factors are associated with the other prevalent risk factors, such as low physical activity (41.7%), high fat food intake (33.3%) and low fruit and vegetable intake (29.6%). Over half of the participants (52.8%) also indicated that they felt stressed most days. Nearly one third of participants were smokers (29.2%).

Table 2 shows the mean systolic and diastolic BP of participants attending the CHAP-EMS during the first five visits. The results show that there was a trend of decreasing BP over time. There was a significant drop in the systolic BP and diastolic by the 3^rd and the 5^th CHAP-EMS visit respectively.

Overall, 41.7% of the participants had elevated BP (either systolic or diastolic, >140/90 respectively) during the initial visit. Of participants previously diagnosed with hypertension, 53.5% had elevated BP. Of participants not previously diagnosed with hypertension, 40.7% had elevated BP. Of those who had elevated BP at the initial visit (n = 27), 41% had BP that was within the normal range by visit 5, and 36% by visit 10 (see Fig. 1). Of those presenting with elevated systolic BP at visit 1 (n = 26), 46% had normalized systolic BP by visit 5 and 36% by visit 10; of those presenting with elevated diastolic BP at visit 1 (n = 8), 86% had normalized diastolic BP at both visits 5 and 10.

Many people who had completed the CANRISK assessment and were at risk of developing diabetes or already had it, also had elevated BP at the initial visit (either systolic or diastolic > 130/80 respectively; see Table 3). Though not statistically significant there was a tendency towards normalization of BP across the ten visits, in people who had a moderate/high CANRISK score.

Participants who had specific risk factors were referred to available resources (exercise programs, food and nutrition resources, Community Care Access Centre (CCAC), social welfare organizations, public health, etc.). Most of the regular participants were referred to and enrolled in the in-house wellness program run by the Municipal authorities, which included an exercise program and scheduled lectures. Participants who were noted to have elevated BP were referred to see their family physicians. Three participants who were previously undiagnosed returned with a diagnosis of hypertension and were subsequently treated. Five participants previously diagnosed with hypertension had their medications adjusted. Furthermore, because of risks identified from the CANRISK assessment, one participant was confirmed to have diabetes by their family physician and two participants had their diabetes medications adjusted. Participants confirmed by the TUG test to be at risk of falls were referred to the CCAC for appropriate assistance, though we were unable to ascertain factual information about what happened subsequently due to confidentiality.

Examining the literature, differing costs could be assumed following a 911 call. There could be an ambulance sent [costing between $240 (19) and $785 (20)] and there is often an emergency department (ED) transport [costing between $259(21) and $627 (22)] with an assessment in the ED [costing $842 (23)]. As per the literature, we assumed that once 911 is called in Ontario, an ED visit is mandatory [19]. Using the above references each paramedic visit and subsequent mandatory hospital ED physician evaluation were estimated to be an average of $1626, ranging from $499 as the lowest estimation of cost and $2254 as highest estimation of cost [19‐23]. A sensitivity analysis demonstrated that the decrease of 20 EMS calls demonstrated by this pilot in one building, could decrease healthcare costs by as little as $9980 [19, 20] or as much as $45,080 per year [19, 20, 22], or by an average of $32,520. Paramedics staffing the CHAP-EMS program were on modified duties (e.g. due to injury), so they were being paid by the Paramedic Services but could not perform a traditional paramedic role due to health-related limitations. Therefore, there was no cost associated with having these paramedics on modified duty staff the CHAP-EMS program. The training for paramedics was provided at no cost for the pilot study. However, the estimated cost for training is nominal since it was a single 3-h training session, which has since been translated into an online training site, which is also free.

Anecdotal narratives emerged from the program that add depth to our understanding of how CHAP-EMS assisted in the prevention of major morbidity for these participants. Some of these stories are related here.

A female participant previously thought to have had acid reflux, but with a history of angina, attended a CHAP-EMS session with acute sternal chest pain, which was diagnosed by the paramedics as severe angina requiring treatment and follow up. She was put in contact with a cardiologist who saw her that day and she was scheduled for balloon angioplasty within 2 weeks where upon her angina disappeared.

Another participant, an elderly male, had refused to call 911 but decided to seek help from the paramedics staffing CHAP-EMS since he was a regular participant. On attending and being assessed it was evident that he had acute chest pain, low BP and was feeling faint. An ambulance was called and he was transported to the ED whereupon he was diagnosed and treated for an acute myocardial infarction.

A female participant was well known to the paramedics and presented with dizziness and when her BP was taken was found to be bradycardic. She was transported by ambulance to the ED where a pacemaker was fitted by cardiology.

Another female participant known to be hypertensive had repeated high BP readings. Upon readings of 180/102 repeatedly one day, paramedics arranged for her to see her family physician that day, who increased her hydrochlorothiazide, added a new medication, and asked to continue being checked by CHAP-EMS.