aja(07)_309.fm Page 205 Thursday, October 23, 2008 3:38 PM
DOI: 10.1111/j.1741-6612.2008.00309.x
Innovations in Aged Care Comprehensive geriatric assessment ‘online’ Blackwell Publishing Asia
Len Gray Academic Unit in Geriatric Medicine, The University of Queensland, Brisbane, Queensland, Australia
Richard Wootton Centre for Online Health, The University of Queensland, Brisbane, Queensland, Australia
This paper describes a system designed to enable comprehensive geriatric assessment to be performed at distant locations. A structured assessment incorporating the interRAI Acute Care assessment tool is administered by a specifically trained nurse assessor onsite. Data are entered and processed by web-based software that incorporates a clinical decision support system. It enables a geriatrician to review and report the assessment online. The assessment and report can be viewed by authorised clinicians inside and outside the hospital via the Internet. The system can also be used to support in person geriatric consultation and whole of episode ward-based geriatric care. Preliminary evaluation suggests the system to be reliable, safe, efficient and appealing to clinicians.
specialist geriatric units. Evidence of the impact of geriatric consultation on patient outcomes is mixed [4]. However, geriatric consultation incorporates important triaging decisions to inpatient geriatric assessment, rehabilitation, long-term institutional care and complex community support programs. Geriatric consultation is delivered by geriatricians and gerontic nurses – alone or in partnership – sometimes with support from other allied health specialists. This process is timeconsuming, particularly when travel is involved, and cannot be delivered in many rural hospitals where geriatric specialist expertise is lacking. In this paper an innovative solution for delivery of CGA is described – ‘online’ comprehensive geriatric assessment. This system has been developed and tested over a 4-year period by the Academic Unit in Geriatric Medicine (AUGM) in association with the Centre for Online Health (COH) at the University of Queensland.
Aims of the service model Key words: assessment, elderly, online, computer, software, clinical decision support.
Introduction The number of older persons in Australia requiring health care will progressively increase over the next few decades. However, the population group that currently constitutes the majority of the health-care workforce will remain virtually stable [1].
An online model for delivery of CGA was developed in response to two primary challenges: • How can specialist expertise be made available in hospitals where it currently does not exist? • How can the dependence on rare and expensive geriatrician time be reduced without compromising quality of care or accuracy of decisions?
The assessment system The specialist workforce in aged care includes geriatricians, gerontic nurses and specialist allied health professionals. The availability of these specialists, and the specialist services within which they work, is already inadequate. Furthermore, the undersupply is mal-distributed, with access considerably worse in provincial cities and rural communities [2]. Older persons by virtue of frailty, limited support and negative attitudes are often not able to travel to specialist centres. The problem of limited aged care specialist supply is a major national and international issue.
The basic system is as follows. The patient is initially assessed by a nurse assessor in person using a standard protocol. The observations are entered onto a web-based software application. The information is then checked for completeness and subsequently reviewed and reported ‘online’ by a geriatrician. Electronic notification enables the completed assessment to be reviewed by relevant clinicians, inside or outside the hospital (Fig. 1). This model is dependent on a range of ‘capabilities’ as follows.
Comprehensive geriatric assessment (CGA) is a core procedure in specialist geriatric care. There is evidence that this process improves functional recovery, reduces morbidity and attenuates demand for long-term institutional care [3]. It is central to geriatric consultation services which are the vehicle for delivering CGA to hospital patients who are not located in
A standardised geriatric assessment instrument incorporating clinical decision support tools The system uses the interRAI Acute Care assessment tool which is designed to support comprehensive geriatric assessment of older patients in the acute hospital setting (see www.interRAI.org). It is a ‘third generation assessment tool’, in that it is purpose-built and part of an integrated multidomain suite of assessment instruments.
Correspondence to: Professor Len Gray, University Department of Medicine, Princess Alexandra Hospital. Email:
[email protected]
It consists of a broad schedule of items designed to characterise the persons’ medical, functional and psychosocial characteristics.
Australasian Journal on Ageing, Vol 27 No 4 December 2008, 205–208 © 2008 The Authors Journal compilation © 2008 ACOTA
205
aja(07)_309.fm Page 206 Thursday, October 23, 2008 3:38 PM
G r a y
Figure 1: The basic schema for ‘online’ assessment.
L ,
W o o t t o n
R
and the patients whom they may review (at individual patient or hospital level). The software processes observations recorded by nurse assessors using the prescribed algorithms provided by interRAI. It compiles a suite of reports available online or in hard copy format. It also incorporates messaging functions that enable notifications to be distributed among staff involved in the case by email or SMS. These functions are universally available, inside and outside of the treating hospital. They can be used to refer patients among internal hospital departments or to community agencies. Clinical information is not distributed directly – rather, upon receipt of a notification, the recipient is authorised to access the patient’s record. The system requires a record of patient consent when a referral is made to an external agency.
There are items pertaining to geriatric syndromes (and risks of acquiring them) with potential data collection points related to the premorbid, admission and discharge periods. It includes a list of working diagnoses and current medications. The interRAI system applies a battery of algorithms to generate scalar measures (for cognition, mood, communication, activities of daily living, instrumental activities of daily living and pain) and risk profiles for common events (such as delirium, falls, discharge to institutional care and readmission). Additional algorithms called ‘clinical assessment protocols’ (CAPs) identify patients most likely to benefit from preventive or restorative interventions (such as fall prevention or need for rehabilitation). The assessment system therefore acts as a clinical decision support system. For less-experienced practitioners it serves the purpose of interpreting basic observations to enhance the diagnostic and evaluation performance of the assessor, whereas for experienced clinicians, it provides a comprehensive problem list and a set of prompts to take specific actions. Nurse assessors with expertise in aged care Assessments are administered by nurses with expertise in aged care and who are specifically trained in the process of performing interRAI assessments and utilising the software. They draw on information derived from patient interview and observation, interview of direct care staff and family members and from the medical record. Observations are recorded on standard data sheets, which consist primarily of categorical data, but with an opportunity to record free text. This approach to collection of clinical data was developed by the interRAI research collaborative. Web-based software enabling record sharing among the consultation team and other stakeholders Data are entered onto a secure, web-based software system developed by the AUGM and COH. This enables data to be entered (and read) by users at any computer connected to the Internet. All users are registered at the individual level and have person-level secure access to the system. ‘Access rights’ are specified at the user level, in relation to tasks that they may perform (assessments, reporting, reading and administration) 206
A training program to enhance the skills of assessors A key ingredient is careful training of staff using the system. Assessors require a 2-day training program to learn the observation and recording conventions and definitions associated with the interRAI Acute Care. Geriatricians require a 3- to 4-hour training session in interpretation of the instrument and familiarisation with software. A 1-hour training session is made available to other staff who wish to ‘read’ and interpret reports compiled by the system.
The model in practice This model has evolved over 4 years, and is now operating in six hospitals in Queensland, Australia. Each site has applied the basic model to suit its particular circumstances. In two cases, there is no visiting geriatrician, and the system is used purely in its online form. Triage decisions and other recommendations are made and implemented without direct discussion with the assessing nurse, although the case may be reviewed at a regular teleconference in the subsequent week. No additional data, beyond the online assessment, is available to the reviewing geriatrician. (In the near future, it is anticipated that online pathology results and radiological reports and images will be available to the geriatrician as the review is conducted.) In complex cases, a telephone call may be made to the nurse assessor for clarification of some aspects, but this occurs in less than 5% of cases. In other settings, it is used to provide a systematic infrastructure to support geriatric consultation, or as an electronic record and decision support system within a ward-based geriatric evaluation and management or rehabilitation ward. Initial assessments may be conducted at the point of admission to a ward, using a case-finding strategy in the emergency department or hospital ward. In some cases a screener is applied to identify cases requiring CGA. In these cases, the admission profile reflects the status of the patient at admission. In hospitals where the system is applied to geriatric consultation on referral, the initial profile will reflect the patient’s status at the time of referral. However, in all cases a premorbid profile is formulated based on the patient’s status before the acute illness. Australasian Journal on Ageing, Vol 27 No 4 December 2008, 205–208 © 2008 The Authors Journal compilation © 2008 ACOTA
aja(07)_309.fm Page 207 Thursday, October 23, 2008 3:38 PM
C o m p r e h e n s i v e
g e r i a t r i c
Two brief case studies are provided to illustrate application of the model. Princess Alexandra Hospital is a 700-bed metropolitan teaching hospital, which includes a 70-bed postacute geriatric and rehabilitation unit. The geriatric consultation service assesses patients who may require postacute care, complex discharge planning or permanent residential care. Nurse assessors provide the initial assessment which is entered online. Geriatricians review assessments online (in some cases preparing ‘draft’ reports before visiting the patient in the ward). After visiting the patient, a report is finalised on a ward-based computer, and printed immediately for insertion in the file. This report provides a comprehensive professional replacement for hand-written progress notes. The assessment is available online for other stakeholders, including staff in the receiving geriatric and rehabilitation wards, Aged Care Assessment Program (ACAP) staff, transition care and allied health staff. Toowoomba Base Hospital is a 300-bed general hospital serving a community of 100 000 people. The hospital includes a 25-bed geriatric and rehabilitation ward. There is no geriatrician in the city. A case finding strategy in the emergency department identifies patients requiring CGA. Screening is supported by a short screener applied in the emergency department which identifies patients who have prior personal activities of daily living dependency, but considerable discretion is used by staff. As the system matured, early case finding became routine in the medical and orthopaedic wards without necessarily requiring formal screening instrumentation. Patients are assessed within 48 hours of admission by a nurse assessor, and reported online by a geriatrician in Brisbane within 4 hours. These assessments provide the basis for triage decisions including transfer to the geriatric ward, which usually occurs within 2 to 4 days. Cases not transferred are followed up by the assessor, and repeat assessments performed when there is a substantial change of status or a key discharge decision is required (e.g. referral to ACAP or transition care). The assessor meets weekly with the geriatrician by tele- or videoconference to discuss case progress. When patients are transferred to the geriatric ward, the geriatrician interacts with patients at the bedside and ward-based staff in case of conference through mobile videoconference. Review and discharge assessments performed by ward-based staff provide an online electronic information system which is viewed by the geriatrician on rounds and during case conferences.
Evidence of efficacy This system is complex and can be evaluated from a wide range of perspectives. There is an ongoing structured evaluation of the system that is broadly based on the framework offered by Fryback for evaluation of radiological systems [5]. The framework uses a six-level hierarchical model of efficacy, ranging from technical accuracy (e.g. psychometric properties of the assessment items), through diagnostic accuracy, diagnostic thinking, therapeutic, patient outcome and societal efficacy Australasian Journal on Ageing, Vol 27 No 4 December 2008, 205–208 © 2008 The Authors Journal compilation © 2008 ACOTA
a s s e s s m e n t
‘ o n l i n e ’
(e.g. improved access to CGA). A range of evaluations has been performed by the AUGM, and others are in progress or planned. These will be the subject of further publications, but some of the preliminary evidence is presented here. An initial study of the interRAI Acute Care systems suggests the item selection to be appropriate with good interrater reliability [6]. A survey of nurse assessors, ward staff, participating geriatricians and referring medical staff found high levels of acceptance among the primary users, and generally positive responses among other staff [7]. These findings are echoed in a three-hospital study which is currently being concluded. A study of timing of assessment documentation showed clear advantages of using an electronic framework, even when assessments were undertaken live [8]. A further study demonstrated that online assessments require approximately one half of the time of live assessments (13.7 minutes vs 22.2 minutes including typing) without taking into account walking or other travel time for live assessments. When the discussion time for case conferences (by video- or teleconference) is taken into account (assuming that this is only required for ‘online’ cases) the time required remains less than for live assessments. This same study found that triage decisions made online agreed with those made live in 61% of cases, and that in no case was an assessment made that would be detrimental to the patient’s interests. This pilot study involved 18 cases. To date, interrater reliability of live assessments has not been measured, essential for precise evaluation of live–online agreement. A larger, multicentre study is planned to further examine diagnostic accuracy and safety. Uptake and sustained use provide additional evidence of acceptance. In all six hospitals where decisions were made at an operational level to use the system, there is sustained and growing use of the system. Over 2000 patients have been assessed. One hospital has been operating for 3 years, one for 2 years and four commenced in the past 12 months. Several time-limited demonstrations have been conducted in private hospitals. In one case, this has translated to an ongoing implementation, but in others the system was closed after external funding was exhausted. However, in all cases, the system was reported to have at least some positive attributes. Up to 10 additional implementations are planned over the next 12 months.
Conclusion This is a system still in evolution, with many aspects undergoing refinement. Perceived weaknesses have been repeatedly addressed through reform of work processes, software adjustments and training. Constant modifications provide a significant challenge in regard to evaluation, but early evidence suggests the approach to be accurate, safe, appealing to clinicians and efficient. All sites have increased access to geriatrician expertise. Hospitals with geriatricians on site have reduced response times and referral rates are increasing. Two cities without geriatricians now have daily access to geriatrician expertise. Similarly, 207
aja(07)_309.fm Page 208 Thursday, October 23, 2008 3:38 PM
G r a y
gerontic nurse specialists are providing new or expanded roles at each site. Overall, the system appears to be meeting its two key objectives of improved access and efficiency.
Acknowledgements
3 4 5 6 7
Key Points • Geriatric assessment can be conducted ‘online’.
W o o t t o n
R
References 1
2
The authors acknowledge the financial assistance provided by the Geriatric Medical Foundation of Queensland and Queensland Health to support this work. The intellectual property including the software utilised within the system is owned by the University of Queensland and has been licensed to Uniquest Pty Ltd for commercialisation.
L ,
8
Australian Bureau of Statistics (ABS). Population Projections, Australia, 2004 to 2101. ABS, 2006. Available from URL: http://www.abs.gov.au/ AUSSTATS/
[email protected]/Lookup/3222.0Main+Features12004%20to% 202101?OpenDocument. Accessed 11 June 2008. Gray L, Moore K, Smith R, Dorevitch M. Supply of inpatient geriatric medical services in Australia. Internal Medical Journal 2007; 37: 270 –273. Stuck AE, Siu AL, Wieland GD, Adams J, Rubenstein LZ. Comprehensive geriatric assessment: A meta-analysis of controlled trials. Lancet 1993; 342: 1032–1036. Gray L. Geriatric consultation: Is there a future? Age and Ageing 2007; 36: 1–2. Fryback DG, Thornbury JR. The efficacy of diagnostic imaging. Medical Decision Making 1991; 11: 88–94. Gray L, Bernabei R, Berg K et al. Standardizing assessment of the elderly in acute care: The interRAI Acute Care. Journal of the American Geriatrics Society 2008; 56: 536 –541. Khateeb M, Gray L. Gazing at the crystal ball: A review of the Geri-Scan. Annual Scientific Meeting of the Australian Society for Geriatric Medicine, 2006; Christchurch, New Zealand, p. A172. Gray L, Vincent R, Martin-Khan M, Varghese P, Wootton R. Processing time for an online geriatric assessment tool. Journal of Telemedicine and Telehealth 2006; 12: 38 – 40.
• Access to specialist expertise for rural hospitals is enabled. • Geriatrician time is utilised more efficiently. • The process appears reliable, efficient and attractive to clinicians.
208
Australasian Journal on Ageing, Vol 27 No 4 December 2008, 205–208 © 2008 The Authors Journal compilation © 2008 ACOTA
