Computer Support System In Nursing
Computerized Decision Support Systems
Although
there is no clear agreement about how to define Computerized Decision Support
Systems (DSSs), most would agree that a DSS can be defined in general as a
computerized system used to aid decision making related to semi-structured
problems.
But some incorrectly include under the umbrella of DSS software that
are not truly DSSS, such as expert systems. While differentiation is fuzzy, in
general a true DSS is a collection of software programs, at the core of which
are mathematical and statistical modeling components which act with real data
to facilitate decision making.
A defining characteristic of DSSS is that they
are proactive. They provide rapid responses to real situations based on real
data, models, and established guidelines.
They are designed to be flexible, and
allow ad hoc queries and easy changing of parameters in order to accommodate
clinician intuition and judgment.
Scope Of DDS
DDS
systems vary in terms of complexity and scope, ranging from simple provision of
integrated reports to use of inferencing methods to determine complex
associations between two pieces of data.
While their goal is to facilitate
effective decision making, they deal with problems that are relatively
unstructured.
For example, such a system might be used to predict how a new
patient care treatment might affect the average duration of patient stay in an
institutional setting.
Goals Of DDS
The
ultimate goal of any DSS is to help clinicians overcome their cognitive
resource limitations for processing and storage as well as problem solving in
an increasingly complex medical environment.
DSSS do this by helping clinicians
to manage information overload in order to properly assess all of the relevant
information and generate systematic and reasonable therapy.
This has the net
effect of facilitating standardization of care. reducing errors, and improving
quality of care.
DDS In Health Care
Healthcare
DSS systems use actual patient data to provide information that can help
clinicians make decisions.
Wyatt and Spiegelhalter (1990) add the requirement
that a medical DSS generate case-specific advice. The use of DSSS in clinical
decision support can be divided into two categories: diagnostic and
therapeutic.
Diagnostic DDS
There
are several types of diagnostic DSSs. First are systems generating differential
diagnoses. Such systems provide lists of possible diagnoses based on given
clinical data.
How- ever, such systems are often problematic as the potential
benefit for the differential diagnosis-generating DSS to inform caregivers
about additional relevant diagnoses can be outweighed by the “noise”
that arises from the presentation of irrelevant or inappropriately ranked
diagnostic choices (Weiner & Pipers , 2000).
Another type of DSS is based
on a rule in/out model. These are used by caregivers to rule in or out a small
set of diagnoses based on a given set of objective clinical signs and symptoms.
They function like a second opinion and have been successful in limited
application (Weiner & Pifer ).
A third type of DSS is used for
computer-aided review of clinical tests such as radiographs or pathology
specimen evaluation ( Alberdi et al., 2000; Peters, 1996). Such systems help
caregivers to interpret results, and have again had success in limited
application.
Therapeutic DDS
Therapeutic
DSSs focus on decision making in point of care treatment. Some focus on medication
dosing, with the goal of reducing errors and complications. Others manage
complex processes such as ventilation and oxygenation (East et al., 1999).
Most
therapeutic DSSS focus on compliance of caregivers with established
quality-of-care guidelines, such as embedding hypertension guidelines within
the hypertensive patient record McAlister, Covvy , Tong, Lee, &
Wigle,1986).
Their goal is to generate, at the point-of-care, patient-specific
evidence-based therapy instructions that can be carried out by different
clinicians with little inter-clinician variability. Individualization of
patient therapy is preserved by these explicit protocols since they are driven
by individual patient data (Morris, 2001).
A good example in nursing of such a
DSS is the Braden System ( Bergstrom , 1997). This is a DSS that guides the
caregiver through risk assessment and then suggests risk-based care tailored to
the specific patient risk-factors based on published guidelines.
However, while
the use of DSSS in therapeutics seems reasonable, research is need that
demonstrate their benefits in terms of outcome measures (Weiner & Pifer ,
2000).
Nursing research in the area of informatics has a history of perhaps 25
years, most of which has been heavily invested in the basic work necessary for
the building of DSS systems .
This basic work includes the development and
identification of classification systems, taxonomies, vocabularies, best
practices, essential data elements, and types of information used in nursing
research and nursing decision making (McCormick & Jones, 1998; Werley ,
Devine, Zorn, Ryan, & Westra , 1991; Benner, 1984).
While nurse
informaticists have also developed circumscribed DDS systems using these
building blocks, research related to the accuracy of the decisions and the
efficacy of these systems in improving outcomes is fairly limited (Johnston ,
Langton , Haynes, & Mathieu, 1994) .
One study was located which tested the
accuracy of a DSS system in using assessment data with a forward chaining
inference engine to identify nursing diagnoses and interventions appropriate to
the patient (Hendrickson & Paganelli , 1994).
A few studies have moved
beyond these basic issues to test the effectiveness of specific DSS systems in
producing nursing decisions that result in better outcomes of care ( Cuddigan ,
Logan, Evans, & Hoesing , 1988; Petrucci et al., 1992 ).
Some have also
moved to development of decision support systems based on established
guidelines (Bowles, 2003).
Future research will likely focus on how DSSS can help
nurses help patients make decisions in scenarios characterized by the need for
careful deliberation about alternatives due to the risk or uncertainty of the
outcomes or the value-laden nature of the decision (O’Connor et al ., 1997).
DDS In Nursing
In
1993, the National Institute of Nursing Research (NINR) constituted an expert
panel on Nursing Informatics. They were charged with setting research
priorities for nursing informatics as part of the National Nursing Research
Agenda.
In carrying out this mandate, the panel identified seven foci for
research, and within each focus, these experts assessed the state of the
science, then identified and prioritized more specific research needs (NINR,
1993).
These foci were:
(a) using data, information, and knowledge to deliver and
manage patient care
(b) defining and describing data and information for
patient care
(c) acquiring and delivering knowledge from and for patient care
(d) investigating new technologies to create tools for patient care
(e)
applying patient-care ergonomics to the patient-nurse-machine interaction
(f)
integrating systems for better patient care
(g) evaluating the effects of
nursing information systems
Similarly, in 2001 lawmak ers provided the Agency
for Health Research and Quality (AHRQ) with $50 million to undertake a major
research initiative investigating the problem of medical errors.
Among funded
projects now under way are four different studies (two in adult and two in
pediatric populations) assessing the impact of using handheld DSSS in ambulatory
care settings (Ortiz & Clancy, 2003).
Health care delivery today is so
complex that it is currently straining the resources of the country, and
multifaceted clinical decisions are being made in an environment of rapidly
escalating intensity.
As DSS systems are developed to produce specific
patient-care protocols that have been validated through using rigorous
methodologies, these systems have the potential to decrease harmful variation
in care, improve clinical decision making, reduce errors, optimize outcomes of
care, and cut health care costs.