Skip to main content

Table 2 Learning health systems with reported outcomes. *Peer-review article

From: Learning health systems using data to drive healthcare improvement and impact: a systematic review

  LHS name Country (scale) Purpose Impact outcomes
1 ePRO Duke cancer clinics LHS [19] United States (Local) To build LHS infrastructure with patient-reported outcomes in EHR in a cancer clinic. Electronic administration of distress screening, provided immediate scoring, & facilitated triage.
Evaluation done on longitudinal impact of a psychosocial care program provided to patients with breast cancer. Significant improvement in distress and despair as measured by self-report questionnaires at 3 and 6 mo *
2 Surgical Care and Outcomes Assessment Program (SCOAP)
[20]
United States (Regional) A peer-to-peer surgeon collaborative that create & track process of care metrics, and interventions to correct under performance. Multisite benchmarking led to decreased percentage of postoperative complications (17.7–9.6%), increased use of imaging, testing, blood glucose checks. *
3 ImproveCareNow Chronic Care Network [21, 22] United States (National) To empower clinicians, researchers, parents and youth to learn and continuously improve care and outcomes for chronic diseases like inflammatory bowel disease. Improvements in the remission rates
Time savings of 7 min per patient visit due to automatic data transfer.
Staff using & benefiting from a learning exchange platform. *
4 NUCATS Institute LHS [22, 23] United States (Regional) To create a central hub supporting clinical and translational research.
Dual-use model warehouse designed to serve both research and clinical needs, integrating healthcare and research.
Greater than a 2.5-fold increase from 2011 to 2013 in data requests by affiliates. *
Created dashboard enabling real-time monitoring of transplant outcomes, replacing a slow manual process.
Wagner et al. 2015 introduced PROMs into routine cancer care, automatically sent to patients prior to appointment, 80% (874) of pts. who read message completed questionnaires.
TOPCAT study recruitment: daily reports identified potentially eligible inpatients (based on free-text, laboratory, imaging, and medication data). Northwestern became top US enrolling site.
Change in care as HFpEF pts. identified in reports are now invited to a specialized clinic.
5 Ottawa Hospital Lung Cancer LHS [27, 28] Canada (Regional) To drive system optimization & innovation in cancer care using community of practice (cop), hub-and-spoke infrastructure, and regional steering committee.
Later, to operationalize LHS thinking, developed the Ottawa
Health Transformation Model (ohtm)
Compliance with provincial evidence-based clinical guidelines improved (20% increase in 2010–11 compared to 2006–07). Other improvements were standardization & implementation of regional perioperative pathways; rectal cancer surgery centralization; increased use of sentinel lymph node biopsies in breast cancer surgery; and decreased positive surgical margin rates in prostate cancer.*
Lung cancer diagnosis now provided to 80% of referrals within target of 28 days. Median patient journey from referral to initial treatment decreased by 48% from 92 to 47 days.
6 PEDS-CHOIR [29] United States (Local) Tertiary care clinic registry to guide research and precision pain medicine in pediatric populations. Captures patient-caregiver PROM/PREM data in real time, to augment clinical decisions and treatment delivery. *
Completion rates increased after staff training, clinic flow enhancements, & conversations with patients, clinicians, and staff highlighting the benefits of personalizing & optimizing patient care over time. Completion adherence increased:
- first pt. survey: 82.4 to 91.7%,
- subsequent pt. surveys 17.3 to 57.6%
7 University of Alabama at Birmingham Hospital, USA [30] United States (Local) Collaboration between healthcare finance leaders & physicians and sharing a vision to create KPIs that pairs patient-centred outcomes with increased efficiency. Surgeons now access KPIs (scorecards), showing performance and benchmarks. Physician committee monitors scorecards & provide feedback.*
(improved quality and savings but no figures provided)
8 ePPOC [31] Australia and New Zealand (National) Chronic pain registry to guide patient clinical care and research, and to establish a benchmarking system to drive multicenter quality improvement. Clinics submit high-quality data describing the demographic and clinical characteristics of patients. Information used in each service to assess & monitor patients and submitted to central coordinating site for analysis, reporting, and benchmarking purposes.*
9 Community Health Applied Research Network (CHARN) [32, 33] Across four states, US (Regional) To pool data, and promote integration of research in health centers and translate it into practice to advance evidence base for improved care in medically underserved patients. Goal1: Fully developed an operational infrastructure to support national Patient Centred Outcomes Research.
Goal2: Fully developed a consensus derived research agenda to guide the networks activities.
Goal3: Fully/partially created processes to develop multimodal proposals, conduct pilot studies and carry out multimodal investigations with support.
Goal4: Not accomplished, transfer of network findings into practice.
Goal5: Fully developed a collaboration for bi-directional education and exchange of ideas, information and values. *
(didn’t achieve Goal 4 translation into practice, but did achieve Goal 5 of bi-directional collaborations)
10 KPNC 30 primary care practices at 13 medical facilities across four counties [34] Northern California, US (Regional) To test the hypothesis that a Pre-Visit Prioritization secure email message could improve visit communication and glycemic control among patients with type 2 diabetes. Improved measures of interaction: more patients reported preparing questions for their visit (72% vs 63%, p = 0.048) and being given treatment choices to consider (81% vs 73%, p = 0.041). Patients in both arms had similar reductions in HbA1c over the 12-month study period (0.56% ± 1.45%) with no significant differences between arms.*
11 Connected Health Cities [35] (plus final report June 2019) UK (Regional) To make better use of information & technology that already exists in health & social care system to improve health and ensure services are more joined-up. Ransom 2018: 6% reduction in average no. of prescription items per person with frailty. Reduction in prescribing costs by £69 – £299 per patient year. Correspondence article
Final report (Jun 2019) - Enabling Data Flows in Greater Manchester Connected Health City: 3 projects described: Opioids, Stroke, and Musculoskeletal epidemiology projects.
At least 20 research projects described on webpage (NOT WITH OUTCOMES YET)
12 PCORnet® [35, 36] US (National) To operationalize the learning health system across several healthcare systems. Provides capacity to conduct transformative clinical research with real-world data, research capabilities, patient partnerships, and broad array of health services researchers.
Disseminating and promoting the uptake of research findings is part of PCORI’s legal mandate to improve the quality and relevance of evidence available to help patients, caregivers, clinicians, employers, insurers, and policymakers make better-informed health decisions.
PCORnet Bariatric Study compared weight measures at 1, 3, and 5 years post surgery (3 types) in 44,978 patients. One type (RYGB) led to greater long term weight loss but also had highest 30-day rate of adverse events. Project results implemented in new shared decision making (SDM) model being evaluated with patients.*
Information on webpage shows research (Arterburn 2018; Toh 2017*) led to a new shared decision making (SDM) model.
PCORI website reports Dissemination & Implementation activities underway for 21 PCORnet projects (Limited Competition Project funded by PCORI). Peer review publications available for some at time of writing:
- Wade 2017* led to Widespread Implementation of a Patient-Centered Online Therapy for Adolescent Traumatic Brain Injury
- Lowenstein et al. 2018* led to Implementing Patient Decision Support for Lung Cancer Screening through Tobacco Quitlines
- Implementing Peer-Driven Care to Patients with Sleep Apnea
13 Swedish Rheumatology Quality Register [39] Sweden (National) To improve the healthcare and treatment of patients with rheumatoid arthritis and other chronic diseases. PROMs and care information are entered as routine clinical practice. Patients benefit by being involved in their own care & records. Physicians benefit by longitudinal overview of each patient including disease activity, disability and treatment. Shared decision making is facilitated. *
Evaluations found patients value system for knowledge it gives them about their changing condition & symptoms over time.
Data used for research purposes: e.g. peer review publications and evidence of safety from drugs (Simard 2011; Neovius 2014).*
14 University of Wisconsin Health LHS [42, 43] United States (Regional) To consistently deliver high value care and support continuous learning and improvement in the health system. System-level performance improvements 2010–2015:
- patient satisfaction improved 0.078 points per month and significant at P > .001
- pneumococcal vaccination increased 62 to 90%, and colorectal cancer screening from 69 to 81%.
- Staff completing formal courses in improvement science tripled between 2012 and 15
- patient & family advisory councils increased by 83% from 90 to 165
Odds of primary care followup doubled (OR 2.0, 95% CI 1.4–2.9). Median time to followup decreased from 71 to 38 days (Bartels 2019). Implementing checklist improved delivery of Family Centered Rounds (Cox 2017).*
Plus on website:
- 52 Toolkits to assist implementation & change initiatives
- - 397 Peer-Reviewed Articles*
15 Wound Care LHS [47, 48] United States (National) To integrate wound care management, quality improvement, & comparative effectiveness research by harnessing real-world data in a purpose-built electronic health record at point of care. Centers submit clinical & quality data, enabling benchmarking across national network. *
It showed patients’ contrasted strikingly with published RCT samples: real-world population are older and sicker, with common comorbidities: e.g. Pressure Ulcers healing rate 40% in 2 RCTs, and 30% in LHS.*
16 Alberta Strategic Clinical Networks [49, 50] Canada (Regional) To support clinically-led, evidence-informed change in Alberta’s health system.
Goals of achieving best outcomes, seeking greatest value for money and engaging multidisciplinary clinicians in all aspects of the work.
Reduced inappropriate use of anti-psychotics in Long-Term Care by 20%. Improved surgical safety through effective implementation of checklist completion (was 40% and rose to 65%) and 3.5% reduction in errors Evidence based stroke care improved: results show reductions in average length of hospital stay by half, improved access to rehab, and better stroke outcomes. (no figures provided) (Noseworthy 2015)
Gramlich 2017 Compliance with the evidence-based guidelines for colorectal surgery recovery *
17 Geisinger Health Systems LHS [51] United States (Regional) To support continuous learning and improvement across the Geisinger Health system. Quality improvement and cost reduction programs mentioned (but no figures provided).*
Website: No show rates of up to 47% reduced to 24% by using routine data to predict no show risk. High risk patients now receive a phone call prior to appointment.
18 Cystic Fibrosis Foundation Patient Registry [52] United States (National) To describe the cystic fibrosis population in the U.S., support epidemiological and clinical research on cystic fibrosis, and improve the quality of cystic fibrosis care. Benchmarking and public transparency.
Case-mix adjusted center results are made public on the web. Patient can access own records, which brings value to encounters for both patients and providers. Centers can access their records to track performance.*
19 Kaiser Permanente Northern California (KPNC) LHS [53, 54] United States (National) To support continuous learning and improvement across the KPNC health systems. Sepsis improvement program, demonstrated significant decreases in mortality: 8.8% in 2011, 9.3% in 2012, and 7.9% in 2013 (P = 0.02). Decreased hospital mortality was observed primarily in patients with a heart failure and/or kidney disease history (P < 0.01). This corresponded to changes in care for patients with heart failure and/or kidney disease.*
20 Distributed Ambulatory Research in Therapeutics Network (DARTNet) [55] United States (National) To transform multi-sourced data into standardized, actionable health information that supports patient care, quality improvement, patient safety and health improvement, and collaborative learning and research. Practices can compare their performance on many measures to each other.
Jenkins 2013 significant improvements in antibiotic use in intervention compared to control practices. The clinical pathways & patient educational material (FROM Jenkins 2013) became part of the DARTNet Learning community and are used by other DARTNet associated clinical organizations.*
Website promotes: learning communities for clinician, practice staff & researchers
21 Optum Labs LHS [56] United States (National) To improve patient care and value in the healthcare system by connecting the generation of evidence with its accelerated translation into practice and its widespread adoption into care delivery.
Partnership between Optum and the Mayo Clinic.
Study compared medication management in 37,501 diabetic patients: cost was less and also longer intervals between insulin for one drug type. Findings translated into guidelines used by care providers.*
Website information & online magazine
Webpage article describes care in ED changed based on anaphylaxis data research: longer observation period & patient education provided with training in injectables.
Over 150 publications linked to website.*
22 Learn From Every Patient LHS [57, 58] United States (Local) To integrate clinical care and research, and to use this knowledge to systematically deliver continual quality improvement in care. Children with CP had 43% reduction in total inpatient days (p = 0.030 vs prior 12mo period), a 27% reduction in inpatient admissions, 30% reduction in emergency department visits (p = 0.001), and 29% reduction in urgent care visits (p = 0.046). Reductions in healthcare costs of 210% (US$7014/child) (Lowes 2017)
Evidence-based change in hip screening procedure by eliminating annual screening x-rays for all CP patients based on LHS data research (Noritz 2018)*
23 IDEA4PS [59, 60] United States (Local) To generate, integrate, and disseminate research throughout the institution and to promulgate those findings for the greater good. Designed, piloted and implemented the “Falls Wheel” into routine practice. A visual display for risk of a patient falling. *
Cardiac monitoring improvements made, including decreased false alarms (18.8 to 9.6%, p < 0.001).*