Since 2008, the Northern Territory Point-of-Care Testing Program has improved patient access to pathology testing for acute and chronic disease management for remote health services.
To evaluate the analytical quality, service delivery, and clinical utility of an expanding remote point-of-care testing network.
Four years (2016–2019) of data on analytical quality, test numbers, and training statistics and 6 months of clinical point-of-care testing data from Abbott i-STATs at remote health services throughout the Northern Territory were analyzed to assess analytical performance, program growth, and clinical utility.
From 2016 to 2019, point-of-care test numbers increased, with chemistry and blood gas testing more than doubling to 8500 and 6000 tests, respectively, troponin I testing almost doubling (to 6000), and international normalized ratio testing plateauing at 8000 tests. Participation in quality control and proficiency testing was high, with quality comparable to laboratory-based analytical goals. A shift toward flexible training and communication modes was noted. An audit of point-of-care test results demonstrated elevated creatinine, associated with chronic kidney disease management, as the most common clinically actionable patient result.
The Northern Territory Point-of-Care Testing Program provides high quality point-of-care testing within remote primary health services for acute and chronic patient management and care. Clinical need, sound analytical performance, flexibility in training provision, and effective support services have facilitated the sustainability of this expanding point-of-care testing model in the remote Northern Territory during the past 11 years.
The vast geographical area and cultural landscape of Australia's Northern Territory pose unique complexities in effective health care delivery. The Northern Territory occupies the northern central part of Australia and covers an area of approximately 1 349 000 km2 (Figure 1).
The Northern Territory has the highest proportion of people living in remote and very remote locations in Australia, with Indigenous Australians comprising 21% to 58% of the population in these areas.1 Indigenous Australians living in the Northern Territory are noted to have a high prevalence of diabetes, kidney disease, cardiovascular disease, and respiratory disease,1–3 which are in part due to the high levels of social disadvantage and extreme remoteness creating barriers to culturally safe health care delivery.2–4
Remote health care is delivered at the community via primary health centers, governed by either the Aboriginal Community Controlled Health Service or the territory government. Day-to-day health service delivery is provided mainly by on-site remote area nurses and Aboriginal health workers. Emergency care is facilitated by telehealth consultation with duty medical officers, with aerial evacuation to the nearest tertiary hospital if required. Pathology services are provided by laboratories in Darwin or Alice Springs or interstate pathology providers, with samples transported from the remote communities via air or road. The extensive distances between the remote health services and the pathology laboratory (Table 1) can delay receipt of pathology results from days to weeks. These delays are suboptimal for chronic care, causing issues with loss to follow-up5 and impracticality in managing acute care situations.6 As such, the remote Northern Territory has niche clinical, cultural, and financial needs suitable to harness the use and benefits of point-of-care testing technology.
Harnessing the Benefits of Point-of-Care Testing: A Historical Overview of the Northern Territory Point-of-Care Testing Program
Since 2008, the Northern Territory Department of Health, in collaboration with the Flinders University International Centre for Point-of-Care Testing, has supported the Northern Territory Point-of-Care Testing Program (NT POCT Program) within the Northern Territory.7,8 The NT POCT Program commenced in August 2008 using the Abbott i-STAT (Abbott Point of Care, Princeton, New Jersey) at 33 remote Northern Territory primary health care services.8
The benefits of point-of-care testing in remote communities led to the rapid expansion of the NT POCT Program. In the first 4 years of the program, more than 500 staff were competent i-STAT point-of-care testing operators, the total number of i-STAT tests performed had increased to 21 250, and the analytical test quality remained satisfactory compared with laboratory benchmarks.7 By 2015, the NT POCT Program had effectively doubled to include all 72 remote health services within the Northern Territory.9 In 2016, 8 remote health services from the Ngaanyatjarra Lands in the Western Desert region of Australia were also enrolled into the NT POCT Program (Figure 1).
In 2016, the clinical effectiveness of international normalized ratio (INR) testing in the NT POCT Program was reported as improved “timeliness, convenience and patient safety while on warfarin therapy.”10 In a further study, medical evacuation data from 6 remote Northern Territory health services identified that the i-STAT point-of-care testing device had enabled evidence-based triaging of 3 acute medical presentations and had ruled out medical evacuation for patients able to be stabilized within the community.11 Furthermore, when used to aid medical decision-making in acutely ill patients, the NT POCT Program was estimated to save the Northern Territory health care system approximately A$21.75 million per annum by the prevention of unnecessary emergency medical retrievals for 3 common acute presentations.6
Sustainable point-of-care testing models rely on continual quality assessment. The NT POCT Program is now in its 12th year of operation with support from the Northern Territory government to 2025. The aim of this paper is to evaluate the analytical quality and service delivery of the NT POCT Program during a recent expansion phase (2016–2019) and investigate the current clinical utility of the point-of-care tests in remote primary health care. This information may guide the development of sustainable regulatory requirements and any future accreditation standards for point-of-care testing in Australia and globally.
Enrolled Health Services
A maximum of 80 remote health services participated in the NT POCT Program from 2016 to 2019. These comprised Department of Health Remote Health Centres in Central Australia and Top End regions, Department of Health Regional Health Centres in Central Australia or Top End (1 center per region), Remote Aboriginal Community Controlled Health Services in Central Australia and Top End regions and Primary Health Care Centres in the Ngaanyatjarra Lands.
The approximate locations of health services enrolled in the NT POCT Program and the largest major public hospitals in the Northern Territory are shown in Figure 1. The actual distance (direct location-to-location) and distance by road between health services in the NT POCT Program and the closest major public hospital for the enrolled jurisdictions are summarized in Table 1.
Training and Competency Assessment
Training resources and methods of i-STAT training and competency assessment for the NT POCT Program have been described elsewhere.7 In 2019, the NT POCT Program training resources were revised, updated, and distributed.
Face-to-face i-STAT training at remote health centers ceased during 2017. Weekly videoconferencing i-STAT training sessions using GoToMeeting (LogMeIn, Inc, Boston, Massachusetts) were the main mode of training delivery. In 2018, an online i-STAT training module was developed, and was available for trainee use from 2019. Operators deemed competent to use the i-STAT for patient testing received a competency certificate for a period of 2 years.
Continuing Education and Support
Monthly feedback reports summarizing i-STAT test utility, quality control, and proficiency testing participation rates and performance and device errors were provided to health center managers and regional district managers. At the end of each month, health services that had not performed the expected number of quality tests were contacted by direct phone call and asked to perform outstanding quality testing.
Continuing i-STAT education was provided to enrolled health services through the quarterly newsletters. In April 2019, a case study12 reported interference of hydroxyurea on i-STAT creatinine measurements in a Northern Territory patient. An audit of the Northern Territory patient care information service in April 2019 demonstrated at least 17 patients were prescribed hydroxyurea and led to the development of a new drug interference reference poster for program participants.12 In addition, a reference poster documenting common preanalytical interferences for i-STAT measurements was developed and issued in 2019.
Participants in the NT POCT Program were supported by a telephone hotline service manned by scientific staff that operated during business hours and/or 24 hours per day, 7 days per week technical support services from Abbott. Hotline, email, and fax communications were logged using an Access database (Microsoft).
The i-STAT devices were remotely supported using a data management application (Central Data Station, Version 5.27a, Abbott) located on a virtual server. From November 2018 to late 2019, enrolled NT POCT Program i-STAT devices were migrated from Central Data Station to an advanced Web-based data manager (InfoHQ, Abbott).
Point-of-Care Tests Performed
From 2016 to 2019, 4 i-STAT test cartridge types were used in the NT POCT Program: (1) CHEM8+, general chemistry test panel; (2) CG4+, blood gas test panel; (3) PT/INR, prothrombin time (PT)/international normalized ratio (INR); and (4) cTnI, cardiac troponin I. A summary of test analytes, blood sample type, sample volume, and time for result for the i-STAT cartridges is shown in Table 2.
Review of Critical Patient Results
From February 2019, patient i-STAT results were reviewed daily, with critical action limits for i-STAT analytes implemented from September 1, 2019, in consultation with the NT POCT Program clinical advisor (Table 3). From September 1 to December 31, 2019, patient i-STAT results reported outside the critical action limits and with accessible Northern Territory Department of Health online clinical notes were tabulated and discussed with the program's clinical advisor as required.
Internal Quality Control
The internal quality control material used from 2016 to 2019 consisted of TriControl (levels 1 and 2, Abbott), PT/INR (level 2, Abbott), and cTnI (level 1, Abbott). Where possible, quality control lot number continuity was maintained for 6 to 12 months. Enrolled health services were required to participate in regular i-STAT quality control testing as scheduled in Table 4. Lot-specific allowable limits of performance for quality control analytes were defined from manufacturer analyte target values and Royal College of Pathologists of Australasia Quality Assurance Program analytical performance specifications for point-of-care testing. Quality control ranges were documented in visual, “traffic-light” result sheets, indicating acceptable (green), caution (orange), and stop (red) quality control result zones. Where quality control results fell outside defined analyte analytical performance specifications (caution or stop zone), participant health services were educated to cease patient testing and contact the i-STAT scientific hotline for technical troubleshooting until quality control results were acceptable.
From 2016 to 2019, between-site quality control imprecision for i-STAT analytes was calculated using data from quality control lot numbers with the highest number of repeated analyses performed in that year. Observed quality control performance was compared with the median imprecision goal for each analyte for the most recent condensed point-of-care testing and point-of-care INR proficiency testing cycle, as reported by the Royal College of Pathologists of Australasia Quality Assurance Program.
From 2016, 6 geographically central remote health services in the NT POCT Program and the program support site (N = 7) were enrolled in the Royal College of Pathologists of Australasia Quality Assurance Program Condensed Point-of-Care and Point-of-Care INR proficiency testing schemes (also known as external quality assurance programs in Australasia). Interim and end-of-cycle proficiency testing reports were reviewed by scientific staff and participant health services notified of proficiency testing performance via monthly program feedback reports.
With the exception of cTnI, the 2019 median between-site proficiency testing imprecision for i-STAT analytes was calculated from the 7 enrolled participants and compared with analyte-specific median imprecision goals, reported in the most recent condensed point-of-care testing and point-of-care INR test cycles by the Royal College of Pathologists of Australasia Quality Assurance Program. However, cartridge prefix changes for cTnI in 2019 (ie, A, D, G, or Y) increased the variability of median cTnI proficiency testing results between linearly paired proficiency testing samples for all Royal College of Pathologists of Australasia Quality Assurance Program i-STAT users (N = 149). There were insufficient data to calculate the cTnI proficiency testing between-site imprecision for NT POCT Program proficiency testing sites using the same cartridge prefix.
Ethics approval for this study was obtained from the Human Research Ethics Committee of Northern Territory Department of Health and Menzies School of Health Research (reference number 2017-2767).
The NT POCT Program and services delivered by the Flinders University International Centre for Point-of-Care Testing are funded by the Northern Territory Department of Health.
Enrolled Health Services and Devices
The total number of enrolled remote health services increased from 78 in 2016 to 80 during 2017–2019. The representation of health services from the Northern Territory and Ngaanyatjarra Lands was relatively stable from 2016 to 2019 (Figure 2).
The total number of enrolled i-STAT devices in 2019 (N = 87) was indicative of 1 device per service in use for the provision of on-site patient testing and/or mobile outreach services by the majority of health services.
From 2016 to 2019, a total of 1037 operators participated in initial i-STAT training; however, the numbers progressively declined from 335 operators (2016) to 190 operators (2019). In the most recent year (2019), the number of operators requiring i-STAT recertification was higher (227 of 417 operators; 54%) than the number of operators participating in i-STAT training for the first time (190 of 417; 46%). The preferred mode of training delivery changed over time, with face-to-face training almost redundant by 2019 (n = 1) and videoconferencing declining from 289 operators (2016) to 154 operators (2019). A self-paced, online training module was used by 35 operators following its availability in August 2019 (Figure 3).
Continuing Education and Support
Between 2016 and 2019, the number of i-STAT hotline telephone calls averaged 105 per month (Figure 4). In addition, from 2019, an average of 132 emails per month and a total of 85 faxes were recorded (Figure 4).
From 2016 to 2019, overall i-STAT patient testing increased 1.7-fold (17 064 tests in 2016 to 29 611 tests in 2019). Use of the CHEM8+ and CG4+ i-STAT cartridges more than doubled from 2016 to 2019, and use of the cTnI cartridges increased steadily to almost double during the same period (Figure 5, A). In contrast, PT/INR cartridge use plateaued between 8000 and 9000 tests per annum from 2017 to 2019 (Figure 5, A).
As a percentage of total patient cartridge use, use of the CHEM8+ (4351 of 17 064 cartridges [25%] in 2016 to 8852 of 29 611 cartridges [30%] in 2019) and CG4+ (2071 of 17 064 cartridges [12%] in 2016 to 6474 of 29 611 cartridges [22%] in 2019) i-STAT cartridges increased from 2016 to 2019. The use of cTnI cartridges remained stable (3478 of 17 064 cartridges [20%] in 2016, 4645 of 23 999 cartridges [19%] in 2017, 5099 of 26 378 cartridges [19%] in 2018, and 5875 of 29 611 cartridges [20%] in 2019) during the 4 years. Use of PT/INR cartridges decreased from 7164 of 17 064 (42%) in 2016 to 8410 of 29 611 (28%) in 2019 (Figure 5, B).
Patient Results Outside Critical Action Limits
Patient results with any measured i-STAT analyte outside defined critical action limits (“flagged” results) reviewed between September 1 and December 31, 2019, for each i-STAT cartridge are reported in Table 5.
Of the 1910 CHEM8+ i-STAT cartridge results evaluated, 226 (12%) had one or more basic chemistry analytes outside the critical action limit (Table 5). Of the 226 flagged CHEM8+ results, creatinine was outside the critical action limit in 84 (37%). This corresponded with data from the patient clinical notes, with 89 reports of chronic kidney disease (5%) associated with the 1910 CHEM8+ i-STAT cartridges evaluated. However, a small proportion of flagged creatinine values were less than 0.2 mg/dL (<18 μmol/L) and corresponded to pediatric patients (Table 5).
Of the 1403 CG4+ i-STAT cartridge results reviewed, 215 (15%) had one or more blood gas analyte outside the critical action limit. Base excess/deficit was flagged in 198 of 215 CG4+ tests (92%), corresponding to respiratory tract infections or respiratory related illness (43 of 1403 cartridges; 3%), general infection or sepsis (41 of 1403 cartridges; 3%) or chronic kidney disease (40 of 1403 cartridges; 3%).
Patient cardiac troponin I was greater than the critical action limit (0.09 ng/mL) in 45 of 1398 results reviewed (3%). The median flagged cardiac troponin I concentration was 0.35 ng/mL (range, 0.09–48.5 ng/mL; N = 45). The median age of patients with cardiac troponin I results outside the clinical action limit was 49 years (range, 27–68 years) and 27 of the 45 patients (60%) were male.
Patient PT/INR testing was greater than the critical action limit (5.0) in 35 of 1825 results evaluated (2%), with a median flagged PT/INR of 5.5. The median age of patients with PT/INR results outside the clinical action limits was 46 years (range, 21–58 years) and 25 of the 35 patients (71%) were female.
Internal Quality Control
The average monthly health service participation rate for scheduled internal i-STAT quality control was high and steady, with 66 of 78 health services (85%) participating in 2016, increasing to 73 of 80 (91%) in 2017, 72 of 80 (90%) in 2018, and 73 of 80 (91%) in 2019 (Figure 6).
The observed between-site imprecision for quality control testing of clinically important analytes from 2016 to 2019 was reported (Table 6). For TriControl and cTnI quality control, the average between-site quality control imprecision at remote health services was lower than or close to the 2019 reported imprecision goal for Australasian laboratories (Table 6). The between-site imprecision for PT/INR (6.5%) was higher than the goal for imprecision (5.1%). The between-site quality control imprecision for i-STAT analytes was stable across the period of 2016 to 2019 (data not shown).
The lowest external proficiency testing participation rate was documented in the first year of proficiency testing enrollment (2016) with a monthly average of 4.40 of the 6 health services submitting all results. Subsequently, the monthly average proficiency testing participation rate improved to 4.98 of 6 (83%) in 2017, was 4.74 of 6 (79%) in 2018, and peaked at 5.64 of 6 health services (94%) submitting all results in 2019 (Figure 6).
With the exception of Pco2 and Po2, the median between-site 2019 proficiency testing imprecision was lower or close to the 2019 reported imprecision goal for all CHEM8+ and CG4+ analytes (Table 7).
The median between-site 2019 proficiency testing imprecision for PT/INR (6.0%) was higher than the goal for imprecision (5.1%).
Since 2008, the NT POCT Program has facilitated high-quality, timely access to essential pathology results for managing acute and chronic disease in remote, predominantly Indigenous communities. The program continues to expand, with the number of enrolled health services throughout the Northern Territory and Ngaanyatjarra Lands at 80 since 2017. According to Australia's Remoteness Structure, the majority (97.5%) of health services within the NT POCT Program are situated in very remote locations, with the remainder classified as outer regional.13 Thus, this program contributes significantly to the Australian government's “Closing the Gap” initiative on addressing health inequality for remote, Indigenous communities.14
A significant expansion of the NT POCT Program in 2015 resulted in upskilling of a large number of remote health care professionals; therefore, it is not surprising the number of i-STAT trainees declined from 2016 to 2019. By 2019, 227 of 417 (54%) of i-STAT operators required recertification training only.
The decrease in face-to-face training sessions seen is due to improvements in videoconference training resources and a demand for more convenient and flexible training options. The decline in videoconferencing numbers seen is most likely due to training saturation, with most permanent remote health staff already competent i-STAT operators. In 2019 a self-paced, online i-STAT training module was introduced with steady participant uptake. Continual challenges with internal information technology firewalls restricting access to online training modules remains a priority for the effective delivery of online training in remote health services.
There is a need for responsive and continuing education and support of remote health professionals performing i-STAT point-of-care testing. Information from patient case studies, hotline communication, and clinical feedback assisted in the development and revision of program resources.
The generally acceptable analytical quality control and proficiency testing results obtained for the NT POCT Program support the i-STAT operator competency interval of 2 years and the training options delivered by the program remain satisfactory.
The number of patient tests performed at remote health services on the i-STAT devices expanded 13-fold from 2008 to 2019, confirming that clinical utility,10,11 stakeholder satisfaction,7,9 and cost-effectiveness6 of i-STAT testing for patient care continues to be recognized in these communities.
Given the high proportion of the remote population with rheumatic heart disease, atrial fibrillation, and heart valve replacement,3,15 it was not surprising that PT/INR testing comprised the majority of i-STAT tests performed in remote health centers. Further highlighted in this study was the importance of INR monitoring in remote communities, with 35 of the 1825 PT/INR test results (2%) greater than the clinical action limit of 5.0, putting these patients at a significantly greater risk of an adverse bleeding event. The overall proportion of PT/INR testing has steadily decreased, which may be, in part, because of the use of newer oral anticoagulants within the last several years.16
The substantial increase in the number and proportion of CG4+ (blood gas) and CHEM8+ tests performed is consistent with the availability of the i-STAT at every remote health service as well as the 2017 Central Australian Rural Practitioners Association Guidelines that incorporated a new clinical protocol for the early recognition of sick or deteriorating patients.17 This protocol required electrolyte and blood gas point-of-care testing for every patient with a remote early-warning score of 3 or more.17 Additionally, the Northern Territory Indigenous population has the highest prevalence of kidney disease in Australia3,18 and high rates of missed dialysis sessions,3,19 thus necessitating high use of the CHEM8+ cartridge to quantify renal function decline and electrolyte imbalance to inform patient triaging. This finding was further supported by creatinine being the most frequent analyte flagging higher than the critical limit for the CHEM8+ cartridge and by the high proportion of chronic kidney disease patients (40 of 1403; 3%) reported with flagged CG4+ results. Missed dialysis sessions are a common presentation seen at health services in this population19 and the ability to perform point-of-care creatinine testing is important in determining whether the patient requires medical evacuation from the remote community.
The high proportion of flagged base excess results during blood gas (CG4+) analysis was consistent with clinical conditions of acid-base imbalance, such as respiratory tract infections,3 long-term respiratory disease, general infection/sepsis, and chronic kidney disease.3 These findings support previous reports of high incidences of avoidable hospitalization due to respiratory disease complications in the Northern Territory20 and the high rate of hospitalization for Indigenous people for care involving dialysis.3
Whilst the number of cardiac troponin I tests steadily increased from 2016 to 2019, the proportion of tests performed in the NT POCT Program remained stable. The median age (49 years) of patients with cardiac troponin I results higher than the clinical action limit highlighted the relatively young age of remote patients experiencing acute coronary syndrome. This age was consistent with the high proportions (36.1% and 56.1%) of self-reported cardiovascular disease in Indigenous people aged 45 to 54 and 55+ years during 2018–2019.3 The median flagged troponin I (0.35 ng/mL) was low compared with both the average flagged troponin (3.07 ng/mL) and the upper flagged troponin range (48.5 ng/mL), confirming that the i-STAT troponin I is used as an early evaluation tool in patients presenting following the onset of chest pain at remote health services.
Despite large increases in patient testing rates and new operators trained, the analytical quality in the NT POCT Program is sustained and, for critical care analytes including sodium, potassium, creatinine glucose, hemoglobin, pH, and lactate, is comparable to laboratory analytical performance goals.9
When consideration is given to the high proportion of laboratories participating (350 laboratories of a total of 432 participants; 81%) in proficiency testing programs compared with remote health participants, it is not surprising that the between-site quality control imprecision for PT/INR (6.5%) in the NT POCT Program was higher than the imprecision goal of 5.1%. This discrepancy may be explained, in part, by the method of PT/INR quality control reconstitution, which involves the accurate transfer of 1.5 mL liquid activator solution to the lyophilized powder. It is reasonable to expect that the accuracy of this task is lower than that achieved by the laboratory when performed by remote health professional staff at remote health services where dispensing volumetric pipettes are not available. Insufficient mixing and consistency of timing associated with PT/INR quality control may also be more variable at remote health services.
In addition to quality control, this study reports high participation and overall satisfactory analytical performance in proficiency testing for the NT POCT Program for the first time. Similar to that reported for the quality control, inaccuracy in INR proficiency testing reconstitution at remote health services (N = 7) is the most probable cause for the INR between-site proficiency testing imprecision being higher than the imprecision goal derived from approximately 80 predominantly laboratory sites. Future consideration for improving the accuracy of reconstituting lyophilized materials at remote health services must be given, as volumetric pipettes are rarely available.
The higher between-site proficiency testing imprecision reported for Po2 and Pco2 when compared with the analytical goals is most likely related to gas exchange between the proficiency testing material and atmosphere within the room, associated with delays in testing at the remote health services. This could potentially be improved by educating staff to perform proficiency testing promptly; however, this is usually reflective of the high workload and emergency-related patient priorities that remote health staff encounter daily.
Australia's Northern Territory continues to face significant challenges in delivering health care. Since 2008, the NT POCT Program has delivered high quality and timely access to pathology results required to effectively manage acute and chronic disease in remote patients. Analytical quality and community accepted point-of-care testing models can be sustained in these environments only if the unique challenges associated with rural and remote primary health care are understood, intensively supported, and managed effectively.
Future studies include using the clinically critical action limits developed in this study for a more intensive audit of critical i-STAT results during 2020. This aims to map the most common critical results across the different regions of the Northern Territory to provide guidance for clinical training and the development of further education resources.
The authors would like to acknowledge the Northern Territory government for providing financial support for the Northern Territory Point-of-Care Testing Program since the program commenced in 2008. In addition, the authors thank and acknowledge all health professional staff working at the rural and remote health services for their valuable contribution to improving the health outcomes of remotely located patients. We also thank all past and present members of the Northern Territory Point-of-Care Testing Program Management Committee and scientific teams for their hard work and contribution to the program's ongoing success.
The Flinders University International Centre for Point-of-Care Testing has received research support from the Northern Territory Department of Health, as acknowledged in the manuscript.
The authors have no relevant financial interest in the products or companies described in this article.