The cost savings of the rural accelerated chest pain pathway for low-risk chest pain in rural general practice: a cost minimisation analysis

Keywords: chest pain, cost minimisation, cost of illness, diagnostic pathways, point-of-care, rural communities, rural general practice, transfer, troponin.

Accelerated diagnostic chest pain pathways (ADP) are able to rapidly identify a cohort of patients that are low-risk for major adverse cardiac events (MACE) who can be discharged safely, reducing hospital admissions and shortening time in emergency departments (ED).^1,2 All urban ED in New Zealand (NZ) have implemented ADP that incorporate high-precision (sensitivity) troponin.¹ Some rural general practices have access to lower precision point-of-care troponin (POC-cTn) and have adopted the rural accelerated chest pain pathway (RACPP) that incorporates POC-cTn.^3,4

A recent prospective evaluation of the RACPP demonstrated that the clinical outcomes (30-day MACE) are equivalent to ADP used in urban emergency departments. The RACPP reduced the requirement for patients with low-risk chest pain (44% of all patients with chest pain) who presented to rural general practice to travel to a distant ED.³ The adoption of the RACPP by rural general practices, therefore, has the potential to reduce costs for both patients and the healthcare system.

This study aimed to estimate the costs associated with assessing patients with low-risk chest pain using the RACPP in rural general practice compared with transporting such patients to a distant ED.

A retrospective cost minimisation analysis was conducted to quantify the costs for patients with chest pain who were assessed with the RACPP and subsequently discharged as low-risk from rural general practice. These were compared with the estimated costs if the patient had been transferred for assessment in a Waikato District Health Board ED. The primary outcome was total costs, but costs were also considered for both the patient and the health system separately.

The Waikato region is in the North Island of New Zealand and has a large rural population. It has a tertiary level hospital based in Hamilton, with four rural hospitals within the region. For many rural towns, acute care is provided by rural general practice, with referral to a rural or urban hospital if required.

All patients aged >18 years who had chest pain and that were assessed as low-risk in rural general practices using the RACPP between 1 June 2018 and 31 December 2019 were invited to participate. Low-risk patients were eligible to be discharged and managed in the community. Full low-risk criteria are described elsewhere.⁵

Patients that did not respond or did not consent were excluded.

Patients were identified using the Pinnacle Primary Healthcare Organisation’s RACPP database and invited by email and follow-up phone call to complete an online survey in June 2020 using the web-based Qualtrics (Provo, UT) platform. Patients who could not complete the online survey were offered the opportunity to answer questions on the phone (with co-investigator, MS).

Information about the patient’s gender, age, ethnicity, address, registered general practitioner and annual income were collected. Patients were also asked to specify the cost of seeing the GP and any prescriptions related to the chest pain episode, time away from home or work, the need for a support person, and whether they used an ambulance to access the general practice during the chest pain episode. The distance from the patient’s location at the time of chest pain symptoms to the referral hospital was calculated using Google Maps. Socioeconomic deprivation was measured using NZ Deprivation Index 2018 (NZDep18) quintiles.^6,7

Costs not provided by the patient were calculated, as described in Box 1. Where the patient did not provide their income, or they were unwaged, the national average income was substituted.⁸ Patients were also asked to estimate the time spent and distance travelled by a support person, if appropriate.

A sensitivity analysis for productivity losses was performed, which included the Jobseeker support benefit, NZ superannuation payment, NZ minimum wage and no productivity losses.^9–11 All costs were in NZ dollars and current as of 31 December 2019.

There are no applicable cost-of-illness studies to guide power calculation, and international studies have shown wide variation in costs incurred for coronary artery disease.^12,13 We estimated that 35 patients will show a difference in the costs incurred at an alpha of 0.05 and with 80% power, where the only difference between the assessment strategies is related to travel (estimated to be NZ$100). Predicting a 50% survey return rate, we anticipated inviting 70 patients to complete the survey.

Frequency (percentage) and median (interquartile range (IQR)) of the costs are presented for categorical and continuous data respectively. The median difference in the total costs, as well as the cost to the patient and system, were calculated. For the differences in costs, 95% confidence intervals (95% CI) were estimated by resampling the data with 1000 bootstrapped samples.

Ethical approval was obtained through the University of Otago’s Human Ethics Committee (H20/033), and consultation was undertaken with the Ngāi Tahu Research Consultation Committee. Consent was obtained from each patient using the Qualtrics (Provo, UT) survey software. Identifying patient information was removed from the data record once the online survey had been completed. All data were encrypted and will be kept securely for a duration of 10 years and then destroyed, as per University of Otago policy.

Sixty-six patients were identified and invited to complete the survey. The response rate was 22.7% (15/66). Most respondents were female (12/15, 80%) and eligible for NZ superannuation (8/15, 53.3%). All were NZ European/Pakeha. Patient demographic information is shown in Table 1. There were no 30-day MACE recorded.

Table 1.  Respondent characteristics for those who participated in the rural accelerated chest pain pathway (RACPP).

The median total cost (combined patient and system costs) attributed to patients who were assessed using the RACPP in general practice was NZD$593 (IQR NZ$550 to NZ$618), whereas the median total cost attributed to patients who would have presented to the emergency department via ambulance was NZ$1896 (IQR NZ$1731 to NZ$2049).

There was a median difference of NZ$1184 (95% CI NZ$1111 to NZ$1468) between the total costs of the two assessment strategies. The median difference in costs for patients was NZ$344 (95% CI NZ$136 to NZ$458), and the median difference in costs for the health system was NZ$977 (95% CI NZ$977 to NZ$977). Table 2 shows the costs incurred by the patient, the health system and the total costs. The difference in costs for individual patients are presented in Supplementary Fig. S1.

Table 2.  Median cost and interquartile range for (a) the patient, (b) the healthcare system and (c) total costs for the assessment of low-risk chest pain.

Lost productivity costs range between NZ$0 (IQR NZ$0 to NZ$60) and NZ$91 (IQR NZ$49 to NZ$111) for patients assessed in general practice with the RACPP and NZ$0 (IQR NZ$0 to NZ$137) and NZ$326 (IQR NZ$195 to NZ$391) for patients assessed in the ED depending on the amount substituted for unknown wage (Supplementary Table S1).

The main finding of this cost minimisation analysis is that there is a potential saving per low-risk chest pain episode (median saving NZD$1184) when the patient is assessed in rural general practice rather than transferred to an ED within the Waikato region. These savings were to both the patient and the healthcare system. Patients had reduced transportation costs and productivity losses. Total cost savings were present regardless of the substituted income level of the patient. The costs of ED assessment and ambulance transfer to the health system exceeded the costs of Primary Options Acute Care funding streams paid to general practices.

This is the first study that we are aware of that provides an estimate of the cost savings of performing acute chest pain assessment in rural general practice compared with performing this same assessment in a distant ED. These findings are consistent with a previous study that has demonstrated that rural patients are faced with considerable extra costs if they are required to travel to urban centres to access ‘free’ outpatient specialist health care.¹⁵

Major limitations to this study included the low survey response rate, despite efforts to email and phone invited participants. The absence of Māori or other ethnic groups in the study cohort limits the widespread applicability of this study to rural NZ. There is a risk of recall bias as participants were asked to recount their historic costs. Fortunately, most of these costs were able to be verified. Asking participants to make educated guesses about potential time away from home or work for themselves and their partners also adds the potential for bias and error; however, this methodology has been successfully used elsewhere.¹⁵

Given the voluntary nature of the study, response bias may have influenced the results. Similarly, ready access to the internet might also have limited responses, although this was mitigated by follow-up phone-calls by investigators. There may also be additional costs that we have not accounted for. These include the costs of major equipment (analysers), the cost of training and on-going quality assurance in either general practice or ED settings. The potential savings from implementing the RACPP will likely outweigh these costs over time. Assumptions made in this study (e.g. regarding fixed costs such as ambulance costs and charges) might also limit the study’s generalisability.

A cost minimisation study is considered a less attractive methodology than other health economic analyses;¹⁴ however, it is a simple and pragmatic methodology that can be justified in this instance, as we have already demonstrated that the clinical outcomes (no recorded 30-day MACE) were comparable to accelerated chest pain pathways used in urban EDs.^1,3

Although limited by sample size, this study suggests that the assessment of chest pain in general practice results in cost savings for low-risk patients compared with transporting the patient to a urban hospital ED. Widespread investment and roll-out of the RACPP in rural (and other) settings has the potential to reduce costs for both patients and the NZ health service.

Supplementary material is available online.

De-identified data will be made available upon reasonable request to the corresponding author.

Tim Stokes is an Editor of The Journal of Primary Health Care, but was blinded from the peer review process for this paper.

This research did not receive any specific funding.