Transcript Exploring South African Physicians` Acceptance of e

Exploring South African Physicians’ Acceptance
of e-Prescribing Technology
Michael Jones, Jason Cohen, Jean-Marie Bancilhon
Background - Written Prescriptions
•
Primary communication between the diagnosing physician and the dispensing pharmacist
•
Difficulty in reading hand written scripts
•
Lost, misplaced, damaged scripts – all requiring a replacement
•
Unnecessary call backs from pharmacists
•
Fraudulent or replicated scripts
•
Prescription errors accounted for over 7000 patient deaths and 1.3 million patient injuries in
the USA in one year *
* (Werner, Nelson & Boehm-Davis, 2012)
E-prescriptions
Electronic prescribing, or e-prescribing, is the use of computers to enter, modify, review, issue
and/or transmit medication prescriptions
•
This electronic practice may be used in place of the regular written prescription, where
instead of a paper form being filled out by the physician, an online form is filled in
•
The resultant e-prescription is then made available to the dispensing pharmacist in
electronic or printed form
Advantages to pharmacists and patients
•
Easy to read and interpret (illegible handwriting, abbreviations, dosage)
•
No more lost or damaged scripts, either electronic or can be reprinted
•
Lowered call backs
•
Costs reduced for the patient through alternative drug options (e.g. Generic)
•
Safer for the patients (up to 86% reduction in adverse affects) *
* (Kaushal et al., 2010; Halamka et al., 2006)
Advantages to physicians
•
Potential time savings from less call backs or redoing lost/refill scripts
•
Decision support through drug incompatibility alerts (drug-drug, drug-age and drug-allergy)
•
Up to date drug information, ICD10 codes, including recalls
•
Patient script history (where available/permissible)
•
Saved scripts for common ailments
e-Prescribing System Example
iMedx (http://www.imedx.com/turborx)
e-Prescription Use
•
Available in a large number of countries worldwide, major strategic focus in the European
Union
•
However, slow diffusion of e-Prescriptions internationally
Disadvantages / Barriers
•
Limited or no time savings in capturing the script
•
Limited use of decision support in simple treatments
•
Cost of the use of the systems could be a barrier
•
Physicians’ trust in the system and its results
•
In some regions, regulatory constraints (physical signature required)
Problem Area
•
e-Prescribing is a high potential eHealth technology
•
A number of issues may limit its wide adoption and use by physicians
•
Physicians are the primary users of e-prescriptions
•
Physicians get less benefits than the patients. How do we get acceptance by physicians
when the main benefits are felt by someone else at a different time?
Theoretical underpinning
•
This study is underpinned by the Unified Theory of Acceptance and Use of Technology
(UTAUT) (Venkatesh, Morris, Davis & Davis, 2003) and its recent extensions (Venkatesh,
Thong & Xu, 2012)
•
This model has been supported by a number of studies in the Health context
–
–
–
Internationally: Various health informatics studies (e.g. Chang, Hwang, Hung, and Li, 2007)
South Africa: Telemedicine acceptance (Cilliers and Flowerday, 2013)
e-Prescribing: Select UTAUT factors used in e-Prescribing acceptance (Wang et al., 2009)
Research Model
Survey Results
•
72 responses out of 639 emailed questionnaires
•
75% of the responding physicians were male, roughly 25% of respondents were between
30-45 years of age and 50% were between 45 and 60 years.
•
48% of the responding physicians reported having done their own research into eprescribing while others reported learning about it from other physicians (28%)
•
Generally there was a high level of acceptance of e-prescribing amongst the responding
physicians (m=4.03)
•
Qualitative and Quantitative findings were generally corroborative
Quantitative findings – Descriptive statistics
Variable
No. of items
λ
Mean
(std dev)
α
ρ
Acceptance
3
4.03 (.98)
.67
.97
-
Performance Expectancy
6
3.86 (.98)
.75
.96
.696***
Effort Expectancy
4
3.95 (.78)
.65
.87
.461***
Social Influence
3
2.85 (1.06)
.94
.97
.314**
Facilitating Conditions
2
4.18 (.67)
.78
.78
.261*
Trust
3
3.56 (.88)
.66
.88
.533***
Price Value
3
3.46 (.89)
.89
.97
.474***
α=Cronbach’s alpha reliability, λ=lowest factor loading, ρ = Spearman correlation
*** p<0.001 ** p<0.01 * p<0.05
Quantitative findings - suggestions
•
Performance expectancy, trust and price value are the three variables most strongly
correlated with acceptance.
•
Performance expectancy: This result suggests that physicians must believe benefits from
use will accrue and will outweigh the time spent prescribing electronically.
•
Trust: Low levels of trust also appear to be a hindrance to acceptance. Physicians must
believe e-prescribing systems will be free of error and capable of delivering on
expectations.
•
Price Value: The benefits will also need to exceed the associated monetary costs
associated with running e-prescribing systems.
Quantitative findings – suggestions continued
•
Effort Expectancy: Any difficulties in using e-prescribing systems are likely to prevent
expected benefits from being realized. e.g. having high learning curves, poor usability and
a cause of frustration
•
Social influence: Correlated moderately with acceptance. Physicians who had heard about
e-prescribing from fellow physicians were somewhat more likely to accept e-prescribing
(m=4.32) than those who reporting learning about it from other sources (m=3.81) or their
own research (m=3.97)
•
Physician computer experience, age and gender played no role in acceptance of eprescribing
Qualitative findings
•
Performance expectancy - concerns about the negative impact:
•
•
•
“E-scripting is impractical … it is best to use traditional methods. Handing a patient a script improves therapeutic
intervention, makes the service personal and in my opinion is better than facing a PC sending scripts out….”
“too time consuming to use during a consultation.”
“I tried to use e-scripting but I still save a lot of time just writing them by hand”
•
Performance expectancy - optimism about the improved performance:
•
“[e-prescribing software X] eliminates all the errors in dosing, and at the same time the software itself has in-built
ICD10 coding ….”
“Makes for easier record keeping and fits into a more comprehensive ePatient Record System”
•
Qualitative findings - continued
•
Effort expectancy – vendors can do much to improve the usability:
•
“The user interface is still quite cumbersome ... I don’t use the program much because of the cumbersome nature
of it … it is not always on line when I am consulting.”
“[e-prescribing software Y] was free but not very user friendly. I used it and was great but my software I use now is
much more friendly and works for me!!”
•
•
Facilitating conditions – absence of internal skills and vendor support caused frustration:
•
•
•
“[The vendor] took more than a month to link me to the system after registering, so I never started using it”
“Need easy step by step explanation setup and how it works and how to use it”
“system works but developer has no interest (as per usual) to make it work well, let alone optimal”
Qualitative findings – additional insights
•
Broader concerns about regulatory and technical environments
•
•
“e-prescribing still requires ink signature to be legal…electronic signature or image should be legalised”
“Interoperability and confidentiality remain the two biggest problems”
Conclusions and Implications
•
Physicians were generally accepting of the technology
•
Performance expectancy found to be the strongest correlate of acceptance. e-Prescribing
systems must therefore be designed to bring direct benefit to the physician in the form of
improved productivity and a more effective prescribing process. E.g., Decision tree access
to medications
•
e-Prescribing is a complex task that requires both software and hardware to be optimally
configured. e-Prescribing systems must be stable, accurate and perform consistently
(quality assurance, reliable infrastructure), and must add value without impacting
negatively on the physician’s ability to interact with patients. The systems must be easy to
use, and unobtrusive (tablet vs. desktop) at the point of care.
Conclusions and Implications continued
•
Acceptance may however be slow if physicians are responsible for the costs of running and
supporting the systems. Financial support and incentives may be important to the
technology’s success.
•
Physician practices often lack necessary technology infrastructure and skills. Vendor
monitoring and outreach are essential to ensure that physicians have up-to-date software
and functional hardware. Vendor support can go a long way to removing the frustration and
barriers to use, but must be delivered in a reliable manner.
•
Legislature: electronic signatures
•
Comprehensive integration into other eHealth systems, and connectivity into dispensing
pharmacy systems
Further research opportunities
•
Consider the inter-relationships and combined effects of the factors on acceptance which
we have not reported on here.
•
Consider the hardware platforms, e.g. tablet vs. desktop, downloaded application vs. web
application.
•
This study’s performance expectancy scale focused mostly on productivity benefits to the
physician. Future work should focus on physician perceptions of benefits e.g. to patient’s
experience.
•
Impact studies could be undertaken to confirm the technology’s potential to improve the
safety of the scripting process.
Questions or Comments
?
Research Method – Instrument and Sample
•
Data was collected via a cross-sectional survey, using a structured online questionnaire,
sent via email
•
All variables were measured using multi-item 5 point Likert scales adopted from the
literature
•
Respondents were provided an opportunity to add qualitative comments
•
A sampling frame was constructed by a local e-prescription application provider and a list
of physicians extracted from a public list of physicians. This provided a combined sample of
639 potential respondents