Centralising and optimising decentralised stroke care systems: a simulation study on short-term costs and effects

The costs associated with stroke care continuum: a systematic review

Stroke, a leading cause of death and long-term disability, has a considerable social and economic impact. It is imperative to investigate stroke-related costs. The main goal was to conduct a systematic literature review on the described costs associated with stroke care continuum to better understand the evolution of the economic burden and logistic challenges. This research used a systematic review method. We performed a search in PubMed/MEDLINE, ClinicalTrial.gov, Cochrane Reviews, and Google Scholar confined to publications from January 2012 to December 2021. Prices were adjusted using consumer price indices of the countries in the studies in the years the costs were incurred to 2021 Euros using the World Bank and purchasing power parity exchange rate in 2020 from the Organization for Economic Co-operation and Development with the XE Currency Data API. The inclusion criteria were all types of publications, including prospective cost studies, retrospective cost studies, database analyses, mathematical models, surveys, and cost-of-illness (COI) studies. Were excluded studies that (a) were not about stroke, (b) were editorials and commentaries, (c) were irrelevant after screening the title and abstract,(d) grey literature and non-academic studies, (e) reported cost indicators outside the scope of the review, (f) economic evaluations (i.e., cost-effectiveness or cost-benefit analyses); and (g) studies not meeting the population inclusion criteria. There may be risk of bias because the effects are dependent on the persons delivering the intervention. The results were synthetized by PRISMA method. A total of 724 potential abstracts were identified of which 25 articles were pulled for further investigation. The articles were classified into the following categories: 1)stroke primary prevention, 2) expenditures related to acute stroke care, 3) expenditures for post-acute strokes, and 4) global average stroke cost. The measured expenditures varied considerably among these studies with a global average cost from €610-€220,822.45. Given the great variability in the costs in different studies, we can conclude that we need to define a common system for assessing the costs of strokes. Possible limitations are related to clinical choices exposed to decision rules that trigger decisions alerts within stroke events in a clinical setting. This flowchart is based on the guidelines for acute ischemic stroke treatment but may not be applicable to all institutions.

Cost-effectiveness of improvement strategies for reperfusion treatments in acute ischemic stroke: a systematic review

BACKGROUND

Reducing delays along the acute stroke pathway significantly improves clinical outcomes for acute ischemic stroke patients eligible for reperfusion treatments. The economic impact of different strategies reducing onset to treatment (OTT) is crucial information for stakeholders in acute stroke management. This systematic review aimed to provide an overview on the cost-effectiveness of several strategies to reduce OTT.

METHODS

A comprehensive literature search was conducted in EMBASE, PubMed, and Web of Science until January 2022. Studies were included if they reported 1/ stroke patients treated with intravenous thrombolysis and/or endovascular thrombectomy, 2/ full economic evaluation, and 3/ strategies to reduce OTT. The Consolidated Health Economic Evaluation Reporting Standards statement was applied to assess the reporting quality.

RESULTS

Twenty studies met the inclusion criteria, of which thirteen were based on cost-utility analysis with the incremental cost-effectiveness ratio per quality-adjusted life year gained as the primary outcome. Studies were performed in twelve countries focusing on four main strategies: educational interventions, organizational models, healthcare delivery infrastructure, and workflow improvements. Sixteen studies showed that the strategies concerning educational interventions, telemedicine between hospitals, mobile stroke units, and workflow improvements, were cost-effective in different settings. The healthcare perspective was predominantly used, and the most common types of models were decision trees, Markov models and simulation models. Overall, fourteen studies were rated as having high reporting quality (79%-94%).

CONCLUSIONS

A wide range of strategies aimed at reducing OTT is cost-effective in acute stroke care treatment. Existing pathways and local characteristics need to be taken along in assessing proposed improvements.

[Organization and costs of stroke care in outpatient settings: Systematic review]

OBJECTIVE

To review the bibliography on stroke costs (ICD-10 code I63) in the field of primary care.

DESIGN

Systematic review.

DATA SOURCES

PubMed/Medline, ClinicalTrials.gov, Cochrane Reviews, EconLit, and Ovid/Embase between 01/01/2012-12/31/2021 with descriptors included in Medical Subject Heading (MeSH).

SELECTION OF STUDIES

Those with a description of the costs of activities carried out in the out-of-hospital setting. Systematic reviews were included; prospective and retrospective observational studies; analysis of databases and total or partial costs of stroke as a disease (COI). Articles were added using the snowball method. The studies were excluded because: a) not specifically related to stroke; b) in editorial or commentary format; c) irrelevant after review of the title and abstract; and d) gray literature and non-academic studies were excluded.

DATA EXTRACTION

They were assigned a level of evidence according to the GRADE levels. Direct and indirect cost data were collected.

RESULTS AND CONCLUSIONS

Thirty studies, of which 14 (46.6%) were related to post-stroke costs and 12 (40%) to cardiovascular prevention costs. The results show that most of them are retrospective analyzes of different databases of short-term hospital care, and do not allow a detailed analysis of the costs by different segments of services. The possibilities for improvement are centered on primary and secondary prevention, selection and pre-hospital transfer, early discharge with support, and social and health care.

Optimising acute stroke pathways through flexible use of bed capacity: a computer modelling study

Background

Optimising capacity along clinical pathways is essential to avoid severe hospital pressure and help ensure best patient outcomes and financial sustainability. Yet, typical approaches, using only average arrival rate and average lengths of stay, are known to underestimate the number of beds required. This study investigates the extent to which averages-based estimates can be complemented by a robust assessment of additional ‘flex capacity’ requirements, to be used at times of peak demand.

Methods

The setting was a major one million resident healthcare system in England, moving towards a centralised stroke pathway. A computer simulation was developed for modelling patient flow along the proposed stroke pathway, accounting for variability in patient arrivals, lengths of stay, and the time taken for transfer processes. The primary outcome measure was flex capacity utilisation over the simulation period.

Results

For the hyper-acute, acute, and rehabilitation units respectively, flex capacities of 45%, 45%, and 36% above the averages-based calculation would be required to ensure that only 1% of stroke presentations find the hyper-acute unit full and have to wait. For each unit some amount of flex capacity would be required approximately 30%, 20%, and 18% of the time respectively.

Conclusions

This study demonstrates the importance of appropriately capturing variability within capacity plans, and provides a practical and economical approach which can complement commonly-used averages-based methods. Results of this study have directly informed the healthcare system’s new configuration of stroke services.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-022-08433-0.

Discrete-Event Simulation to Model the Thrombolysis Process for Acute Ischemic Stroke Patients at Urban and Rural Hospitals

Background: Effective treatment with tissue plasminogen activator (tPA) critically relies on rapid treatment. Door-to-needle time (DNT) is a key measure of hospital efficiency linked to patient outcomes. Numerous changes can reduce DNT, but they are difficult to trial and implement. Discrete-event simulation (DES) provides a way to model and determine the impact of process improvements.

Methods: A conceptual framework was developed to illustrate the thrombolysis process; allowing for treatment processes to be replicated using a DES model developed in ARENA. Activity time duration distributions from three sites (one urban and two rural) were used. Five scenarios, three process changes, and two reductions in activity durations, were simulated and tested. Scenarios were tested individually and in combinations. The primary outcome measure is median DNT. The study goal is to determine the largest improvement in DNT at each site.

Results: Administration of tPA in the imaging area resulted in the largest median DNT reduction for Site 1 and Site 2 for individual test scenarios (12.6%, 95% CI 12.4–12.8%, and 8.2%, 95% CI 7.5–9.0%, respectively). Ensuring that patients arriving via emergency medical services (EMS) remain on the EMS stretcher to imaging resulted in the largest median DNT improvement for Site 3 (9.2%, 95% CI 7.9–10.5%). Reducing both the treatment decision time and tPA preparation time by 35% resulted in a 11.0% (95% CI 10.0–12.0%) maximum reduction in median DNT. The lowest median and 90^th percentile DNTs were achieved by combining all test scenarios, with a maximum reduction of 26.7% (95% CI 24.5–28.9%) and 17.1% (95% CI 12.5–21.7%), respectively.

Conclusions: The detailed conceptual framework clarifies the intra-hospital logistics of the thrombolysis process. The most significant median DNT improvement at rural hospitals resulted from ensuring patients arriving via EMS remain on the EMS stretcher to imaging, while urban sites benefit more from administering tPA in the imaging area. Reducing the durations of activities on the critical path will provide further DNT improvements. Significant DNT improvements are achievable in urban and rural settings by combining process changes with reducing activity durations.

Impact of ongoing centralization of acute stroke care from "drip and ship" into "direct-to-mothership" model in a Dutch urban area

When acute stroke care is organised using a "drip-and-ship" model, patients receive immediate treatment at the nearest primary stroke centre followed by transfer to a comprehensive stroke centre (CSC). When stroke care is further centralised into the "direct-to-mothership" model, patients with stroke symptoms are immediately brought to a CSC to further reduce treatment times and enhance stroke outcomes. We investigated the effects of the ongoing centralization in a Dutch urban setting on treatment times of patients with confirmed ischemic stroke in a 4-year period. Next, in a non-randomized controlled trial, we assessed treatment times of patients with suspected ischemic stroke, and treatment times of patients with neurologic disorders other than suspected ischemic stroke, before and after the intervention in the CSC and the decentralized hospitals, the intervention being the change from "drip and ship" into "direct-to-mothership". Our findings provide support for the ongoing centralization of acute stroke care in urban areas. Treatment times for patients with ischemic stroke decreased significantly, potentially improving functional outcomes. Improvements in treatment times for patients with suspected ischemic stroke were achieved without negative side effects for self-referrals with stroke symptoms and patients with other neurological disorders.

A Neurodisparity Index of Nationwide Access to Neurological Health Care in Northern Ireland

Nationwide disparities in managing neurological patients have rarely been reported. We compared neurological health care between the population who reside in a Health and Social Care Trust with a tertiary neuroscience center and those living in the four non-tertiary center Trusts in Northern Ireland. Using the tertiary center Trust population as reference, neurodisparity indices (NDIs) defined as the number of treated patients resident in each Trust per 100,000 residents compared to the same ratio in the tertiary center Trust for a fixed time period. NDIs were calculated for four neurological pathways—intravenous thrombolysis (iv-tPA) and mechanical thrombectomy (MT) for acute ischemic stroke (AIS), disease modifying treatment (DMT) in multiple sclerosis (MS) and admissions to a tertiary neurology ward. Neurological management was recorded in 3,026 patients. Patients resident in the tertiary center Trust were more likely to receive AIS treatments (iv-tPA and MT) and access to the neurology ward (p < 0.001) than patients residing in other Trusts. DMT use for patients with MS was higher in two non-tertiary center Trusts than in the tertiary center Trust. There was a geographical gradient for MT for AIS patients and ward admissions. Averaged NDIs for non-tertiary center Trusts were: 0.48 (95%CI 0.32–0.71) for patient admissions to the tertiary neurology ward, 0.50 (95%CI 0.38–0.66) for MT in AIS patients, 0.78 (95%CI 0.67–0.92) for iv-tPA in AIS patients, and 1.11 (95%CI 0.99–1.26) for DMT use in MS patients. There are important neurodisparities in Northern Ireland, particularly for MT and tertiary ward admissions. Neurologists and health service planners should be aware that geography and time-dependent management of neurological patients worsen neurodisparities.

Pathway Design for Acute Stroke Care in the Era of Endovascular Thrombectomy: A Critical Overview of Optimization Efforts

The efficacy of intravenous thrombolysis and endovascular thrombectomy (EVT) for acute ischemic stroke is highly time dependent. Optimal organization of acute stroke care is therefore important to reduce treatment delays but has become more complex after the introduction of EVT as regular treatment for large vessel occlusions. There is no singular optimal organizational model that can be generalized to different geographic regions worldwide. Current dominant organizational models for EVT include the drip-and-ship- and mothership model. Guidelines recommend routing of suspected patients with stroke to the nearest intravenous thrombolysis capable facility; however, the choice of routing to a certain model should depend on regional stroke service organization and individual patient characteristics. In general, design approaches for organizing stroke care are required, in which 2 key strategies could be considered. The first entails the identification of interventions within existing organizational models for optimizing timely delivery of intravenous thrombolysis and/or EVT. This includes adaptive patient routing toward a comprehensive stroke center, which focuses particularly on prehospital triage tools; bringing intravenous thrombolysis or EVT to the location of the patient; and expediting services and processes along the stroke pathway. The second strategy is to develop analytical or simulation model-based approaches enabling the design and evaluation of organizational models before their implementation. Organizational models for acute stroke care need to take regional and patient characteristics into account and can most efficiently be assessed and optimized through the application of model-based approaches.

Centralising acute stroke care within clinical practice in the Netherlands: lower bounds of the causal impact

Background

Authors in previous studies demonstrated that centralising acute stroke care is associated with an increased chance of timely Intra-Venous Thrombolysis (IVT) and lower costs compared to care at community hospitals. In this study we estimated the lower bound of the causal impact of centralising IVT on health and cost outcomes within clinical practice in the Northern Netherlands.

Methods

We used observational data from 267 and 780 patients in a centralised and decentralised system, respectively. The original dataset was linked to the hospital information systems. Literature on healthcare costs and Quality of Life (QoL) values up to 3 months post-stroke was searched to complete the input. We used Synthetic Control Methods (SCM) to counter selection bias. Differences in SCM outcomes included 95% Confidence Intervals (CI). To deal with unobserved heterogeneity we focused on recently developed methods to obtain the lower bounds of the causal impact.

Results

Using SCM to assess centralising acute stroke 3 months post-stroke revealed healthcare savings of $US 1735 (CI, 505 to 2966) while gaining 0.03 (CI, − 0.01 to 0.73) QoL per patient. The corresponding lower bounds of the causal impact are $US 1581 and 0.01. The dominant effect remained stable in the deterministic sensitivity analyses with $US 1360 (CI, 476 to 2244) as the most conservative estimate.

Conclusions

In this study we showed that a centralised system for acute stroke care appeared both cost-saving and yielded better health outcomes. The results are highly relevant for policy makers, as this is the first study to address the issues of selection and unobserved heterogeneity in the evaluation of centralising acute stroke care, hence presenting causal estimates for budget decisions.

Optimising acute stroke care organisation: a simulation study to assess the potential to increase intravenous thrombolysis rates and patient gains

OBJECTIVES

To assess potential increases in intravenous thrombolysis (IVT) rates given particular interventions in the stroke care pathway.

DESIGN

Simulation modelling was used to compare the performance of the current pathway, best practices based on literature review and an optimised model.

SETTING

Four hospitals located in the North of the Netherlands, as part of a centralised organisational model.

PARTICIPANTS

Ischaemic stroke patients prospectively ascertained from February to August 2010.

INTERVENTION

The interventions investigated included efforts aimed at patient response and mode of referral, prehospital triage and intrahospital delays.

PRIMARY AND SECONDARY OUTCOME MEASURES

The primary outcome measure was thrombolysis utilisation. Secondary measures were onset-treatment time (OTT) and the proportion of patients with excellent functional outcome (modified Rankin scale (mRS) 0-1) at 90 days.

RESULTS

Of 280 patients with ischaemic stroke, 125 (44.6%) arrived at the hospital within 4.5 hours, and 61 (21.8%) received IVT. The largest improvements in IVT treatment rates, OTT and the proportion of patients with mRS scores of 0-1 can be expected when patient response is limited to 15 min (IVT rate +5.8%; OTT -6 min; excellent mRS scores +0.2%), door-to-needle time to 20 min (IVT rate +4.8%; OTT -28 min; excellent mRS scores+3.2%) and 911 calls are increased to 60% (IVT rate +2.9%; OTT -2 min; excellent mRS scores+0.2%). The combined implementation of all potential best practices could increase IVT rates by 19.7% and reduce OTT by 56 min.

CONCLUSIONS

Improving IVT rates to well above 30% appears possible if all known best practices are implemented.

Rationale and design for studying organisation of care for intra-arterial thrombectomy in the Netherlands: simulation modelling study

INTRODUCTION

The introduction of intra-arterial thrombectomy (IAT) challenges acute stroke care organisations to provide fast access to acute stroke therapies. Parameters of pathway performance include distances to primary and comprehensive stroke centres (CSCs), time to treatment and availability of ambulance services. Further expansion of IAT centres may increase treatment rates yet could affect efficient use of resources and quality of care due to lower treatment volume. The aim was to study the organisation of care and patient logistics of IAT for patients with ischaemic stroke in the Netherlands.

METHODS AND ANALYSES

Using a simulation modelling approach, we will quantify performance of 16 primary and CSCs offering IAT in the Netherlands. Patient data concerning both prehospital and intrahospital pathway logistics will be collected and used as input for model validation. A previously validated simulation model for intravenous thrombolysis (IVT) patients will be expanded with data of the MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands) Registry and trials performed in the Collaboration for New Treatments in Acute Stroke consortium to represent patient logistics, time delays and outcomes in IAT patients. Simulation experiments aim to assess effectiveness and efficiency of alternative network topologies, that is, IAT with or without IVT at the nearest primary stroke centre (PSC) versus centralised care at a CSC. Primary outcomes are IAT treatment rates and clinical outcome according to the modified Rankin Scale. Secondary outcomes include onset-to-treatment time and resource use. Mann-Whitney U and Fisher's exact tests will be used to estimate differences for continuous and categorical variables. Model and parameter uncertainty will be tested using sensitivity analyses.

ETHICS AND DISSEMINATION

This will be the first study to examine the organisation of acute stroke care for IAT delivery on a national scale using discrete event simulation. There are no ethics or safety concerns regarding the dissemination of information, which includes publication in peer-reviewed journals and (inter)national conference presentations.

TRIAL REGISTRATION NUMBER

ISRCTN99503308, ISRCTN76741621, ISRCTN19922220, ISRCTN80619088, NCT03608423; Pre-results.

Can clinical audits be enhanced by pathway simulation and machine learning? An example from the acute stroke pathway

OBJECTIVE

To evaluate the application of clinical pathway simulation in machine learning, using clinical audit data, in order to identify key drivers for improving use and speed of thrombolysis at individual hospitals.

DESIGN

Computer simulation modelling and machine learning.

SETTING

Seven acute stroke units.

PARTICIPANTS

Anonymised clinical audit data for 7864 patients.

RESULTS

Three factors were pivotal in governing thrombolysis use: (1) the proportion of patients with a known stroke onset time (range 44%-73%), (2) pathway speed (for patients arriving within 4 hours of onset: per-hospital median arrival-to-scan ranged from 11 to 56 min; median scan-to-thrombolysis ranged from 21 to 44 min) and (3) predisposition to use thrombolysis (thrombolysis use ranged from 31% to 52% for patients with stroke scanned with 30 min left to administer thrombolysis). A pathway simulation model could predict the potential benefit of improving individual stages of the clinical pathway speed, whereas a machine learning model could predict the benefit of 'exporting' clinical decision making from one hospital to another, while allowing for differences in patient population between hospitals. By applying pathway simulation and machine learning together, we found a realistic ceiling of 15%-25% use of thrombolysis across different hospitals and, in the seven hospitals studied, a realistic opportunity to double the number of patients with no significant disability that may be attributed to thrombolysis.

CONCLUSIONS

National clinical audit may be enhanced by a combination of pathway simulation and machine learning, which best allows for an understanding of key levers for improvement in hyperacute stroke pathways, allowing for differences between local patient populations. These models, based on standard clinical audit data, may be applied at scale while providing results at individual hospital level. The models facilitate understanding of variation and levers for improvement in stroke pathways, and help set realistic targets tailored to local populations.