A conceptual treadmill: the need for'middle ground'in clinical ...

Journal of Advanced Nursing, 1999, 30(5), 1222±1229
Nursing theory and concept development or analysis
A conceptual treadmill: the need for `middle
ground' in clinical decision making theory
in nursing
Carl Thompson DPhil BA(Hons) RGN
Research Fellow, Centre for Evidence-Based Nursing,
Genesis 6, Science Park,
University of York, York, YO10 5DQ, England
E-mail: cat4@york.ac.uk
Accepted for publication 21 December 1998
THOMPSON C. (1999) Journal of Advanced Nursing 30(5), 1222±1229
A conceptual treadmill: the need for `middle ground' in clinical decision
making theory in nursing
This paper explores the two predominant theoretical approaches to the process
of nurse decision making prevalent within the nursing research literature:
systematic-positivistic approaches as exempli®ed by information processing
theory, and the intuitive-humanistic approach of Patricia Benner. The two
approaches' strengths and weaknesses are explored and as a result a third
theoretical stance is proffered: the idea of a cognitive continuum. According to
this approach the systematic and intuitive theoretical camps occupy polar
positions at either end of a continuum as opposed to separate theoretical planes.
The methodological and professional bene®ts of adopting such a stance are also
brie¯y outlined.
Keywords: clinical decision making, information processing, intuitive knowing,
cognitive continuum, theoretical discussion, methodological development,
nursing knowledge
INTRODUCTION: THE SEMANTICS OF NURSE
operational face of nursing knowledge is the one adopted
DECISION MAKING
by this paper. However, as will be demonstrated later, the
operationalization of knowledge is no monotonic or linear
In relation to the examination of decision making in
equation along the lines of:
nursing a number of expressions are used by authors to
describe what is, in essence, the same phenomenon:
[scienti®c] knowledge + clinical challenge + registered practi-
decisions taken by nurses relating directly to issues of
tioner ˆ uniformly optimal decision
nursing diagnosis or intervention in clinical settings.
An alternative way of viewing the subtle differences in
Clinical decision making is the most common term used
descriptions of the same process is to recognize that terms
(Ford 1979, Field 1987, Luker & Kenrick 1992) but other
such as clinical diagnosis represent the process of
terms utilized include clinical judgement (Benner &
decision making and also the outcome of this process
Tanner 1987, Itano 1989), clinical inference (Hammond
(Carpenito 1983, Gordon 1987). This paper explores theory
1964), clinical reasoning (Grobe et al. 1991) and diag-
on decision making as a process. The paper, and the study
nostic reasoning (Carnevali et al. 1984, Radwin 1990). To
with which it is associated, do not seek to judge the quality
all intents and purposes these terms are interchangeable
of decisions reached, or whether the decisions are better or
given they describe a single process, namely, the `opera-
worse for different nurses using different decision making
tionalisation of nursing knowledge' (Luker & Kenrick 1992
models. There are two reasons why deploying such an
p. 458). The stance that nursing decisions represent the
approach would be methodologically unwise.
1222
Ó 1999 Blackwell Science Ltd

---

Nursing theory and concept development or analysis
Clinical decision making theory
First, studies seeking to examine the outcomes of deci-
databases is represented by a four-stage process (Radwin
sions often depend on the implicit assumption that `good'
1990, Hamers et al. 1994).
equals `accurate' (Tanner et al. 1987, Hamers et al. 1994).
First, the clinician takes part in a patient encounter and
This is problematic for nursing as we do not yet know what
gathers preliminary clinical information about the patient
constitutes an `accurate' nursing diagnosis or intervention-
(also called the cue acquisition stage). This information
decision. For example, would an accurate decision be one
can also be gathered prior to patient encounter.
which most nurses (however ill-informed) would take?
Following this the clinician generates initial, tentative,
One which is best supported by scienti®cally rigorous and
hypotheses. These are related to gathered data and short-
generalizable research ®ndings? Or one which best
term memory-based cues. The number of hypotheses
balances the needs of the nurse, available resources and
generated is generally estimated as being limited to
the end-user? All of these could be taken legitimately as
between four and six.
`accurate' decisions but each could conceivably be very
The third stage (interpretation) involves the clinician
different both in terms of process and outcome.
interpreting the cues gathered during the acquisition stage
This ¯aw is compounded by the second reason for
and classifying them as either con®rming, refuting or not
rejecting a comparative approach, namely, the lack of
contributing to the initial hypotheses generated.
consensus regarding nursing diagnoses themselves.
Using this classi®cation, the ®nal evaluatory stage
Nursing lacks an internationally recognized database of
involves the clinician weighing up the pros and cons of
nursing diagnoses such as that used in medicine in
each decision alternative and choosing the one favoured
specialities such as psychiatry. Moreover, little, if
by the preponderance of the evidence.
anything, is known about the correlation between
This basic sequential hypothetico-deductive model has
information, the cues used to guide decisions, and the
been used as the basis for more elaborate schemata; but
diagnoses or decisions reached. Clearly then without
despite increasing complexity the basic stages remain. For
greater knowledge of the outcomes of decisions and a
example, Carnevali (1984) describes a seven-stage process
demonstrable (or at least visible) degree of consensus
of diagnostic reasoning in nursing which simply breaks
amongst practitioners regarding nursing diagnoses such
down the basic four-stage model described thus far:
an approach would be ill-advised. However, these limita-
1 Exposure to pre encounter data.
tions do not prevent the exploration of the processes
2 Entry to the data search ®eld and shaping the direction
involved in decision making.
of data gathering.
Despite the lack of linguistic homogeneity, a variety of
3 Coalescing of cues into clusters or `chunks'.
models of decision making in nursing, both normative and
4 Activating possible diagnostic explanations (hypoth-
descriptive, have been advanced. These models fall into
eses).
two discrete theoretical categories: the systematic-posi-
5 Hypothesis and data directed search of the data ®eld.
tivist and the intuitive-humanist.
6 Testing diagnostic hypothesis for goodness of ®t.
7 Diagnosis.
THE SYSTEMATIC-POSITIVIST STANCE
ON DECISION MAKING
A number of studies have attempted to go beyond the
somewhat vague assertion that nurses simply weigh up
The dominant explanatory stances on nurse decision
the `pros and cons' of decision alternatives and have
making until the 1980s were models portraying decision
tested the hypothesis that nurses draw on formal or
making as a hypethetico-deductive rational process based
informal (i.e. known or estimated) probability estimates of
around theory derived from the ®eld of cognitive psycho-
`diagnostic ®t' as well as contextual data in making
logy. A number of in¯uential studies took the issue of
clinical decisions. Speci®cally, a number of authors have
nurse decision making and explicitly or implicitly sought
utilized Bayesian and/or probability theory to examine the
to examine it from the perspective of what will be termed
clinical decision making of nurses (Hammond et al. 1967,
the Information Processing Model (Cianfrani 1984).
Aspinall 1979, Panniers & Kellogg-Walker 1994).
The information processing model
Bayesian logic in nurses' decision making?
The central assumption behind the model is that the
Bayes' theorem, expressed mathematically, argues that:
human decision system can be separated into two compo-
Prob…H/E† ˆ Prob…H† Â Prob…E/H†aProb…E†
nents: short- and long-term memory. Short-term memory
houses the stimuli information required to `unlock' factual
The signi®cance of Bayes' theorem to a discussion of
(semantic) and experimental (episodic) knowledge
clinical decision making comes when this equation is
stored in long-term memory (Carnevali et al. 1984, Hamers
unpacked. Simply stated, Bayes' theorem argues that
et al. 1994). The interface between these two cognitive
people hold degrees of belief in relation to scienti®c
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229
1223

---

C. Thompson
theories or outcomes (or indeed any phenomenon). More-
making rather than describing the reality of clinical
over, these degrees of belief will be adjusted in response to
practice. This is a criticism which is levelled at the
the presentation of new probability-evidence (Papineau
information processing model generally; namely, that the
1996), to the extent that the practitioner considers E likely
linear sequential implications of the model are not
given H, but unlikely otherwise. For example, suppose
observed in practice. Nurses frequently overlap stages in
that H is the hypothesis held by the practitioner and E is
the process and change their order (Lauri 1982, Jenkins
some new evidence relating to the hypothesis then the
1985, Corcoran 1986,). Fischoff and Beyth-Marom (1983)
theorem dictates that upon discovering the evidence
present a useful overview of the theory of Bayesian logic
the practitioner will adjust their degree of belief in the
in relation to systematic perspectives on cognition and
hypothesis in line with the right hand side of the equation.
also point out the limitations of the theory in the `real
Expressed alternatively, if the evidence (E) is very
world' of clinical practice. This lack of descriptive `®t' is
surprising or enlightening (for example, that the use of
the basis for the competing, intuitive-humanist, alterna-
compression bandaging signi®cantly decreases healing
tive model of clinical decision making.
time for venous leg ulcers) but is in line with the theory
adopted (H Ð that compression bandaging appears to be
THE INTUITIVE-HUMANIST STANCE
bene®cial in the patients that you treat for leg ulceration)
ON DECISION MAKING
then it should make you increase your belief in (H). The
corollary being that if the evidence is no more likely given
Just as there is a lack of consensus over the terms used to
H than it would be according to any other theory then the
describe decision making, the lack of consensus
presence of E provides no more support for H and beliefs
surrounding descriptions of the role of intuition in clinical
should be adjusted accordingly (Papineau 1996 pp. 295±
decision making mitigates against an easy summary of the
296). The currency of Bayes' theorem therefore is proba-
relevant literature in this area. The various approaches to
bility relating to nursing evidence and hypotheses.
de®ning intuition include:
Hammond et al. (1967), in what was probably the ®rst
· `Understanding without a rationale' (Benner & Tanner
signi®cant strand of nurse decision making research,
1987 p. 23).
examined the ways in which six nurses revised their
· `A perception of possibilities, meanings and relation-
diagnoses of patients' conditions as new data were gath-
ships by way of insight' (Gerrity 1987 p. 63).
ered and presented. These revised hypotheses and diag-
· `Knowledge of a fact or truth, as a whole; immediate
noses were then compared with calculated probabilities of
possession of knowledge; and knowledge independent
the conditions. Whilst consistently reviewing their
of the linear reasoning process' (Rew & Barron 1987
hypotheses, nurses tended to be `cognitively cautious' in
p. 60).
their revisions. That is, they were not as revisionist as the
· `Immediate knowing of something without the
Bayesian model of decision making would suggest.
conscious use of reason' (Schrader & Fischer 1987 p. 45).
Aspinall (1979) took this approach further and found
· `[A] ¼ process whereby the nurse knows something
that a structured decision modelling tool (a decision tree)
about a patient that cannot be verbalized, that is
based on calculated probabilities helped nurses to reach
verbalized with dif®culty or for which the source of
`correct' diagnostic decisions. Panniers and Kellogg-
knowledge cannot be determined. (Young 1987 p. 52).
Walker (1994), however, using a similar approach (though
different research methods) found that nurses' intuitive
The author most attributed with developing the intu-
decisions and those promoted by a tool based on calcu-
itive model and the distinction between theoretical know-
lated probabilities were signi®cantly different. Speci®-
ledge and experiential knowledge is Patricia Benner
cally, in relation to wound dressings there was only a 35%
(1984). Her work has been hugely in¯uential in the
level of agreement between qualitative judgement and
preparation of trainee nurses (Luker & Kenrick 1992) and
quantitative evidence-led prescriptions (Panniers &
offers a useful theoretical counter to the empiricism
Kellogg-Walker 1994), although it is worth noting that in
associated with the information processing model. The
the absence of treatment outcome data (and associated
basic thrust of all intuitive-humanist models is that
probabilities) the quantitatively `correct' decisions in this
intuitive judgement distinguishes the expert from the
study were not necessarily the optimal ones in terms of
novice, with the expert no longer relying on analytical
clinical effectiveness. Instead, the work focused on the
principles to connect their understanding of the situation
probabilities of variables such as nurses' knowledge of the
to appropriate action. Nursing appears intuitive to the
product, patient comfort, risk of adverse effects and costs.
outside observer and feels internalized within the practi-
It is clear that the Bayesian annex of the systematic-
tioner; clinical decisions are the result of an almost
positivist strand of decision making is more prescriptive
unconscious level of cognition (Hamers et al. 1994).
than descriptive. Bayesian models and studies examining
In common with others, McKenna makes the analytical
their utility offer the potential for improving decision
distinction between `know how' and `know that'
1224
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229

---

Nursing theory and concept development or analysis
Clinical decision making theory
knowledge. `Know how' knowledge is that which is skills
communicate something which is intangible and which
based, rooted in intuition and often relates to the `art' of
the practitioner is unable to express. Given that the
nursing. `Know that' knowledge, on the other hand, is
intuitive model exempli®ed by Benner and her acolytes
grounded in theory and empirical research and often
relies on experiential knowledge as the basis for `knowing'
classi®ed as responsible for the `science' of nursing
as opposed to the science of communicable research
(McKenna 1997). For McKenna, Benner's work contributes
®ndings, it is dif®cult to imagine a scenario where
®rmly to the `know how' knowledge base of nursing or
nursing's knowledge base becomes a shared resource
nursing as art. For Benner herself, however, such a classi-
available to all practitioners equally. One could argue
®cation is unforthcoming, as for her nursing constitutes a
that the intangible `character' of nursing is often passed
practice rather than an art (Bishop & Scudder 1997).
from expert to novice through observation and a form of
The primary distinction between systematic-rational
physical apprenticeship, but this still begs the question,
and intuitive approaches lies in their respective motiva-
how does the novice know whether their interpretation of
tional loci. In the systematic approach of the information
the intangible is an appropriate one? This is particularly
processing model the prime motivators in any decision are
so when one considers that certain kinds of clinical
its related task-features: the number of cues, the task
experience and the decisions that accompany them are
complexity, etc. In the intuitive approach of Benner and
something of a `scarce resource' for students, with all the
Dreyfus and Dreyfus (1986) Ð on which Benner's work is
accompanying limitations on opportunities for knowledge
based Ð the shaping force in any decision is the individual
reinforcement that follow.
making it. Good decisions are those made intuitively by
Systematic-rational, hypothetico-deductive, models are
professionals with expertise, such expertise representing
not without their problems in this regard, however.
the end-point of a ®ve-stage sequential transformation
Whilst
undoubtedly
promoting
communicability
from novice to expert:
(through transparency) in the decision making process,
the use of knowledge, and the reproduction of that
1 Novice: those with no experience of situations in which
knowledge, the process itself may not be that relevant if,
they are expected to perform and who ®nd themselves
as appears to be the case, it does not `®t' with the reality
governed by context-free rules as guides to action.
of clinical practice. Lauri and Salantera
2 Advanced beginners: those who demonstrate margin-
È (1995), using a
factor-analytical approach, found evidence that both
ally acceptable performance and have amassed enough
Benner's intuitive model and the hypothetico-deductive
experience to recognize recurring meaning in the
approach of information processing had a degree of
situations they are involved in.
analytical utility in explaining the decision making of
3 Competent: those who see their actions as part of a
nurses. The implication was that both have something to
longer-term plan which helps achieve ef®ciency and
offer and that neither one offers a unitary solution to
organization in work.
explaining decision making in the complex arena of
4 Pro®cient: the practitioner begins to perceive things as
practice. The importance of this pluralistic explanation
a whole with speedy alterations to the long-term plans
is highlighted later in the paper when the idea of a
when expected normal patterns of care do not present
cognitive continuum is introduced.
themselves.
5 Expert: one who has no reliance on guiding rules or
maxims and who has an intuitive grasp of situations;
Simpli®cation
only falling back on hypethetico-deductive logic when
Linked to this realization is the problem of simpli®cation
a new or unexpected challenge arises.
or reductionism. If the information processing model is
failing to capture all the variables involved in decision
THE STRENGTHS AND LIMITATIONS OF
making and clinical diagnosis (McGuire 1985) and at the
THEORETICAL DECISION MAKING MODELS
same time communicating this `incomplete' picture to
other practitioners in the form of scienti®c `evidence' then
Both the systematic-positivist and the intuitive-humanist
nursing's knowledge base will continue to develop in an
stance have a number of strengths and limitations. These
ad hoc manner with signi®cant gaps in the total picture.
can be separated into four themes: communicability;
The intuitive model of expertise and decision making at
simpli®cation; context speci®city; and applicability.
least allows for the complexity of decisions allied to
health care provision and recognizes that health is more
Communicability
than the sum of its constituent parts. More importantly
intuitive expertise-based approaches recognize that
Knowing can only become shared knowledge when it is
nursing as a verb is more than the sum of a series of
communicated to others (McKenna 1997) and herein lies
physical, social and spiritual interventions carried out on
the problem for intuitive models: it is almost impossible to
the patient. However, if the profession relies on holistic
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229
1225

---

C. Thompson
`intuitive' forms of knowing in its practitioners as the
Rempushki (1985) found that whilst decision making was
basis for practice then, whilst undoubtedly allowing for
grounded in the acquisition of data it was far from the
the complexity of the work, nursing decisions will remain
linear progression assumed by the hypothetico-deductive
an opaque activity unable to be in¯uenced by anyone
model generally. Smith (1988) found that intuition or `gut
other than nurses, thereby representing perhaps the ulti-
instinct' was often combined with `objective' data as well
mate form of occupational closure!
as subjective variables such as nurse experience and
familiarity with the patient.
Context speci®city
A DECISION MAKING CONTINUUM?
Crow and colleagues (1995) point to the importance of
practice context to decision making Ð although they
There is little to convince this researcher that either the
prefer the more formal term domain-speci®c knowledge
humanistic-intuitive approach or the systematic-rational
structures. For Crow, contexts such as occupation and
approach offers a solely convincing basis for explaining
clinical specialty are seen as signi®cant determinants of
decision making and by implication the information used
decision making. These domain [context] speci®c know-
as the basis for nursing decisions. A more appropriate
ledge structures:
stance might be to recognize that each has something
to offer and that in their theoretically `pure' states
¼ specify what action to take and can be best described as ways of
they represent ideal-typical frameworks for analysis.
thinking about problems met in everyday practice. For example,
Certainly, White et al. (1992) found in their study that
nurses experienced in looking after patients with diabetes may
whilst hypothetico-deductive models were applied the
judge whether the patient's condition is stable very differently
end-result was a decision making model which possessed
from nurses experienced in looking after patients with a myocar-
the characteristics of both models:
dial infarction.
(Crow et al. 1995 p. 208)
[The study ®ndings] ¼ indicate that the hypothetico-deductive
Crow points to three studies (Prescott et al. 1989, Marks
process of clinical decision making was applied¼ the differ-
et al. 1991, Javacone & Dostal 1992) as evidence of the
ence¼ in the time spent working through the simulations and
importance of such domains to the decisions made by
in the amount of subjective data acquired also is consistent
nurses. The importance of domain or context, however, is
with evolving cognitive models which indicate that ef®ciency
not treated equally by the two camps of clinical decision
in clustering information develops with experience in a given
making theory. The systematic-positivist approach to
setting or with speci®c presentations. This supports the ®nd-
decision making can be criticized because it includes an
ings of Benner that expert nurses¼ move ahead more quickly
implicit assumption that judgement is the result of a
and on the basis of less subjective data than novice nurses.
unitary generic process used by all clinicians at all times
(White et al. 1992 p. 157)
(Berner 1984). The intuitive-humanist approach can be
Similarly, in their work on learning to use scienti®c
criticized for the opposite reason; namely, because of its
knowledge in education and practice settings (in a British
axiom that each clinical challenge is unique, and the
context) Eraut and colleagues (1995) argue that systematic-
processes and inputs used subjective, then it is almost
rational approaches alone fail to explain advanced levels
entirely context-speci®c and transferability of `intuition'
of clinical performance. However, they add the caveat that
between different practitioners becomes impossible
the intuitive models of Benner and the Dreyfusses are
(Radwin 1990).
often idealized rather than researched and that profes-
sional deliberations are usually mixtures of intuitive and
Applicability
analytical processes.
With these points in mind it is appropriate to view both
Whilst both the systematic-positivist model of information
models as ideal typical end-points on a continuum of
processing and the intuitive-humanist approach are
clinical decision making, particularly as there is so little
presented in the literature as descriptive models (i.e. they
empirical material to draw on, and that which is presented
describe how decisions are made rather than how deci-
throws up contradictory themes and mixed messages. This
sions ought to be made) there appears little to convince
view of a `cognitive continuum' in relation to clinical
the author that either offers a unitary solution to the
decision making is supported by Hamm (1988). His analysis
problems of researching clinical decision making in
is based ®rmly in medicine, but the key points apply equally
nursing. A number of commentators highlight the fact
to an analysis of nursing practice. For Hamm, practitioner
that most studies are characterized by decision making
cognition is neither purely intuitive nor purely analytical,
models which represent a `middle way' Ð combinations
rather it is commonly located at some point in between
of intuition, explicit data gathering, and tangible expla-
(Hamm 1988 p. 82). For example, the expert community
nation and intangible `knowing'. For example, Philips and
practitioner will, in the delivery of care:
1226
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229

---

Nursing theory and concept development or analysis
Clinical decision making theory
· Exhibit elegant and logical use of diagnostic and
objective and quantitative then analysis is
decision making skills and expected probabilities in
commonly a feature.
selecting products for use on a patients with wounds
3.3 Time available Ð the shorter the available time the
(rational-analytic thought).
more likely that intuitive approaches will be
· Know intuitively when to refer a patient to another
adopted.
member of the multidisciplinary health care team after
However, where the cognitive continuum theory
just a few introductory questions and an assessment
departs from the `cold' rationality of the information
when faced with a patient with a leg ulcer (intuitive
processing model is in its encompassing of the variables
cognition).
of power, social structure and individual knowledge
· In the course of teaching juniors, steer them in the
(Hamm 1988 p. 84). As a framework for researching
assessment of the presenting features of a given patient
nurse decision making the theory allows for the realiza-
using their own judgements and perceptions as a
tion that multidisciplinary team members will accept
framework (a combination of both Ð quasi rational
analytical thinking from people who are broadly
cognition).
perceived as competent, able to eliminate uncertainty,
According to the cognitive continuum theory the major
and familiar: often those `experts' in senior clinical
determinants of whether a practitioner utilizes a rational
positions. Similarly practitioners may reject intuitive
or intuitive approach to decision making are primarily
solutions from `junior' colleagues where analytical
determined by the position of the task on a continuum
reasoning cannot be demonstrated. Clearly, this suggests
which has three dimensions (Hamm 1988 p. 83):
that variables such as an individual's position in the
structures and hierarchies in a work environment will
1 Complexity of task structure.
exert some sway on the organizing cognitive principles
1.1 Number of cues Ð when presented with lots of
available for them to deploy.
information a practitioner will probably utilize an
Similarly, the relationship between individual know-
intuitive approach.
ledge and cognitive modes is signi®cant; as Hamm points
1.2 Redundancy of cues Ð the more cues help in the
out, `the ability of a task thinker to induce a mode of
prediction of the presence of other cues then the
cognition depends [also] on what the thinker knows'
more likely that intuitive cognition will be used.
(Hamm 1988 p. 84). For example, if a practitioner does not
1.3 The nature of an organizing principle Ð if a simple
know that there are sound scienti®c principles behind the
`averaging' approach to combining information is
selection of wound care products, then a wound-based
known to be more accurate then intuitive thought
`task feature' will not encourage analytical cognition in
is likely to be a feature. If it is known that a
dealing with wound data.
complicated approach to combining evidence
The fact that issues of social structure and levels of
produces more accurate answers then this will
knowledge variation amongst practitioners can be incor-
induce an analytical approach.
porated into any analysis is an attractive aspect of the
2 Ambiguity of the task
theory. From a researcher's perspective this allows nurse
2.1 Whether an organizing principle exists Ð if an
decision making to be brought into analytical and meth-
organizing principle exists then the practitioner is
odological frameworks which view nursing as a form of
more likely to be analytical.
social action, with the attendant emphasis on questions of
2.2 Familiarity of the task Ð unfamiliarity induces an
culture, values, interests and power that accompany such
intuitive approach as the practitioner has not had
perspectives.
time to develop more complicated ways of dealing
with cue information.
2.3 The potential for accuracy Ð if a particular
CONCLUSION
approach to assessment is known to be accurate
(even if only perceived as such) then it is more
The nursing literature commonly separates decision
likely to be used as the basis for analysis. For
making into one of two theoretical camps: the system-
example, universal assessment scales for pressure
atic-positivist approach (as typi®ed by information
sore assessment.
processing theory), and the intuitive-humanist stance
3 Nature of the presentation of the task
of Benner and the Dreyfusses. However, neither of these
3.1 Task decomposition Ð if the task leads to the need
two positions offer a unitary theory able to reconcile the
to address related sub-tasks then analytic modes of
apparently different worlds of normative theory and
thought will be used.
clinical reality. An alternative approach is to accept
3.2 The ways in which information is presented Ð if
that, whilst conceptually distinctive, the two approaches
visual information is used then intuition is
occupy the same theoretical plane. Speci®cally,
induced. If the information is presented as
the systematic-positivistic stance of the information
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229
1227

---

C. Thompson
processing-based approaches and the intuitive-human-
istic stance of Benner represent poles at either end of a
References
cognitive continuum.
Aspinall M. (1979) Use of a decision tree to improve accuracy of
The promotion of the two approaches as poles on a
diagnosis. Nursing Research 28, 182±185.
continuum rather than diametrically opposed entities is
Benner P. (1984) From Novice to Expert: Excellence and Power
not simply some theoretical `opt out' or a simple method-
in Clinical Nursing Practice. Addison-Wesley, Reading,
ological `middle-way'. It is a solution which recognizes
Massachusetts.
the diversity of individual cognitive strategy and, more
Benner P. & Tanner C. (1987) Clinical judgement: how expert
importantly, lends itself to the theoretical development
nurses use intuition. American Journal of Nursing 87(1), 23±31.
and empirical testing of decision support systems which
Berner E. (1984) Paradigms and problem solving: a literature
review. Journal of Medical Education 59, 625±633.
assist the clinician in reaching decisions able to cope with
Bishop A.H. & Scudder J.R. (1997) Nursing as a practice rather
the plurality of modern clinical practice. This is an
than an art or a science. Nursing Outlook 45(2), 82±85.
important point as, in respect of the Benner/Dreyfus
Carnevali D.L., Mitchell P.H., Woods N.F. & Tanner C.A. (1984)
models, such supporting empirical enquiry is not a
Diagnostic Reasoning in Nursing. Lippincot, Philadelphia.
dominant feature of the literature. Such quantitative
Carpenito L.J. (1983) Nursing Diagnosis. Applications To Clinical
evidence of the explanatory as well as normative power
Practice. Lippincot, Philadelphia.
of the humanistic-intuitive approach to decision making
Cianfrani K.L. (1984) The in¯uence of amounts and relevance of
for nursing would greatly add to the utility of the theory
data on identifying health problems. In Classi®cation of
for the profession.
Nursing Diagnoses: Proceedings Of The Fifth National Confer-
Whilst not wishing to introduce the quantitative/qual-
ence (Kim M.J., McFarland G.K., McLane A.M. eds), Mosby,
itative evidence debate into the paper at such a late stage,
St Louis, pp. 150±161.
Corcoran S.A. (1986) Task complexity and nursing expertise as
it is worth noting that the decisions nurses make, the
factors in nursing decision making, Nursing Research 35(2),
processes behind them and the outcomes they produce are
107±112.
intricately linked to nursing's professionalization `project'
Crow R., Chase J. & Lamond D. (1995) The cognitive component of
Ð i.e. the broader struggle to be seen as a professional
nursing assessment: an analysis. Journal of Advanced Nursing
occupational group. Good quality, generalizable and
22, 206±212.
analytically transparent evidence of the applicability of
Dreyfus H.L. & Dreyfus S.E. (1986) Mind Over Machine: The
Benner's ideas to nursing practice would do much to
Power of Human Intuition and Expertise in the Era of the
convince those concerned with nurses' decision making of
Computer. The Free Press, New York.
the value of this vital member of the health care team,
Eraut M., Alderton J., Boylan A. & Wraight A. (1995) Learning to
particularly as the qualitative studies often used as a
Use Scienti®c Knowledge in Education and Practice Settings:
means of `demonstrating' the theory can be criticized for
An Evaluation of the Contribution of the Biological Behavioural
and Social Sciences to Pre-Registration Nursing and Midwifery
further mystifying professional practice and ultimately
Programs. Research Report Series Number 3. English National
serving as literary tools supporting attempts at occupa-
Board for Nursing, Midwifery and Health Visiting, London.
tional closure. By acknowledging that unitary decision
Field P.A. (1987) The impact of nursing theory on the clinical
theories are limited in their methodological usefulness to
decision making process. Journal of Advanced Nursing 12,
the researcher of nursing practice and decisions, then
563±571.
more elegant Ð and more importantly, transparent Ð
Fischoff B. & Beyth-Marom R. (1983) Hypothesis evaluation from
research solutions can be advanced. This has to be to the
a Bayesian perspective. Psychological Review 90, 239±60.
bene®t of all and not just those who are intuitively `in the
Ford J.A.G., Trygstad-Durland L.N. & Nelms B.C. (1979) Applied
know'.
Decision Making For Nurses. Mosby, St Louis.
Gerrity P. (1987) Perception in nursing: the value of intuition.
Holistic Nursing Practice 1(3), 63±71.
Acknowledgements
Gordon M. (1987) Nursing Diagnosis, Process And Application.
McGraw Hill, New York.
This paper is derived from a much larger review of the
Grobe S.J., Drew J.A. & Fonteyn M.E. (1991) A descriptive analysis
literature on nurse decision making carried out as part of a
of experienced nurses' clinical reasoning during a planning
National Health Service Research and Development
task. Research in Nursing and Health 14, 305±314.
project examining nurses' use of information in clinical
Hamers J.P.H., Huijer Abu Saad H. & Halfens R.J.G. (1994)
decision making being carried out at the Centre for
Diagnostic process and decision making in nursing: a literature
review. Journal of Professional Nursing 10(3), 154±163.
Evidence Based Nursing, University of York. Particular
Hamm R.M. (1988) Clinical intuition and clinical analysis:
thanks go to Professors David Thompson and Trevor
expertise and the cognitive continuum. In Professional Judge-
Sheldon, Dr Nicky Cullum and Dorothy McCaughan for
ment: A Reader in Clinical Decision Making (Dowey J. &
comments on earlier drafts of this paper. Results from the
Elstein A. eds), Cambridge University Press, Cambridge,
project will be available in April 2000.
pp. 78±109.
1228
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229

---

Nursing theory and concept development or analysis
Clinical decision making theory
Hammond K.R. (1964) An approach to the study of clinical
Panniers T.L. & Kellogg-Walker E. (1994) A decision analytic
inference in nursing: part II. Nursing Research 13(4), 315±319.
approach to clinical nursing. Nursing Research 43(4),
Hammond K., Kelly K., Scheider R. & Vancini M. (1967) Clinical
245±249.
inference in nursing: revising judgements. Nursing Research
Papineau D. (1996) The philosophy of science. In The Blackwell
16, 38±45.
Companion to Philosophy (Bunnin N. & Tsui-James E.P. eds),
Itano J.K. (1989) A comparison of the clinical judgement process
Blackwell, Oxford pp. 290±325.
in experienced registered nurses and student nurses. Journal of
Philips L. & Rempushki V. (1985) A decision making model for
Nursing Education 28(3), 120±126.
diagnosing and intervening in elder abuse and neglect. Nursing
Jacavone J. & Dostal M. (1992) A descriptive study of nursing
Research 34(3), 134±139.
judgement in the assessment and management of cardiac pain.
Prescott P.A., Soeken K.L. & Ryan J.W. (1989) Measuring patient
Advances in Nursing Science 15, 54±63.
intensity: a reliability study. Evaluation and the Health Profes-
Jenkins H.M. (1985) A research tool for measuring perceptions of
sions 12(3), 255±269.
clinical decision making. Journal of Professional Nursing 1(4),
Radwin L.E. (1990) Research on diagnostic reasoning in nursing.
221±229.
Nursing Diagnosis 1(2), 70±77.
Lauri S. (1982) Development of the nursing process through
Rew L. & Barron E. (1987) Intuition: a neglected hallmark of
action research. Journal of Advanced Nursing 7, 301±307.
nursing knowledge. Advances in Nursing Science 10(1), 49±62.
Lauri S. & SalanteraÈ S. (1995) Decision making models of Finnish
Schrader B. & Fischer D. (1987) Using intuitive knowledge in the
nurses and public health nurses. Journal of Advanced Nursing
neonatal intensive care nursery. Holistic Nursing Practice 1(3),
21, 520±527.
45±51.
Luker K.A. & Kenrick M. (1992) An exploratory study of the
Smith S. (1988) An analysis of the phenomenon of deterioration
sources of in¯uence on the clinical decisions of community
in the critically ill. Image: The Journal of Nursing Scholarship
nurses. Journal of Advanced Nursing 17, 457±466.
20(1), 12±15.
Marks R.J., Simons R.S., Blizzard R.A. & Browne D.R.G. (1991)
Tanner C.A., Padrick K.P., Westfall U.A. & Putzier D.J. (1987)
Predicting outcome in intensive therapy units Ð a comparison
Diagnostic reasoning strategies of nurses and nursing students.
of apache ii with subjective assessments. Intensive Care Medi-
Nursing Research 36(6), 58±363.
cine 17, 159±163.
White J.E., Nativio D.G., Kobert S.N. & Engburg S.J. (1992) Content
McGuire C. (1985) Medical problem solving: a critique of the
and process in clinical decision making by nurse practitioners.
literature. Journal of Medical Education 60, 587±595.
Image: The Journal of Nursing Scholarship 24(2), 153±158.
McKenna H. (1997) Nursing Theories And Models. Routledge,
Young C. (1987) Intuition and nursing process. Holistic Nursing
London.
Practice 1(3), 52±62.
Ó 1999 Blackwell Science Ltd, Journal of Advanced Nursing, 30(5), 1222±1229
1229

---