You are on page 1of 11

ISSU ES I N CLINICA L NUR SIN G

Assessment of nursing management and utilization of nursing resources


with the RAFAELA patient classification system case study from the
general wards of one central hospital
Anna-Kaisa Rainio

MSc, RN

Development Manager, Vasa Central Hospital, Vasa, Finland

Arto E Ohinmaa

PhD

Associate Professor Department of Public Health Sciences, University of Alberta, Edmonton, Alberta, Canada

Submitted for publication: 28 January 2004


Accepted for publication: 2 November 2004

Correspondence:
Anna-Kaisa Rainio
Vasa Central Hospital
Hietalahdenkatu 2-4
SF-65130 Vaasa
Finland
Telephone: 35863231725
E-mail: anna-kaisa.rainio@vshp.fi

674

R A I N I O A - K & O H I N M A A A E ( 2 0 0 5 ) Journal of Clinical Nursing 14, 674684


Assessment of nursing management and utilization of nursing resources with the
RAFAELA patient classification system case study from the general wards of one
central hospital
Background. RAFAELA is a new Finnish PCS, which is used in several University
Hospitals and Central Hospitals and has aroused considerable interest in hospitals
in Europe.
Aims and objectives. The aim of the research is firstly to assess the feasibility of the
RAFAELA Patient Classification System (PCS) in nursing staff management and,
secondly, whether it can be seen as the transferring of nursing resources between
wards according to the information received from nursing care intensity classification.
Methods. The material was received from the Central Hospitals 12 general wards
between 2000 and 2001. The RAFAELA PCS consists of three different measures:
a system measuring patient care intensity, a system recording daily nursing resources,
and a system measuring the optimal nursing care intensity/nurse situation. The data
were analysed in proportion to the labour costs of nursing work and, from that, we
calculated the employers loss (a situation below the optimal level) and savings
(a situation above the optimal level) per ward as both costs and the number of nurses.
Results. In 2000 the wards had on average 77 days below the optimal level and
106 days above it. In 2001 the wards had on average 71 days below the optimal level
and 129 above it. Converting all these days to monetary and personnel resources the
employer lost 307,745 or 9.84 nurses and saved 369,080 or 11.80 nurses in total in
2000. In 2001 the employer lost in total 242,143 or 7.58 nurses and saved 457,615
or 14.32 nurses. During the time period of the research nursing resources seemed not
have been transferred between wards.
Conclusions. RAFAELA PCS is applicable to the allocation of nursing resources but its
possibilities have not been entirely used in the researched hospital. The management of
nursing work should actively use the information received in nursing care intensity
classification and plan and implement the transferring of nursing resources in order to
ensure the quality of patient care.
Relevance to clinical practice. Information on which units resources should be allocated to is needed in the planning of staff resources of the whole hospital. More
resources do not solve the managerial problem of the right allocation of resources.

 2005 Blackwell Publishing Ltd

Issues in clinical nursing

Assessment of nursing management and utilization of nursing resources

If resources are placed wrongly, the problems of daily staff management and cost
control continue.
Key words: classification, cost, management, nursing, patient, resources

Introduction
Personnel planning and the optimal allocation of nursing
resources as well as effective management represent an
internationally important subject of health care research.
Literature on the productivity of hospitals (Fare et al. 1997,
Linna 1998, Linna & Hakkinen 1999) shows that there are
clear cost efficiency and productivity differences between
hospitals and wards. These are largely due to their different
staff costs in relation to their output such as bed-days and
outpatient visits. The management of staff resources in
hospitals and their wards requires tools which take into
consideration both the amount and quality of nursing staff
and patients and are able to react to changes in them.
The nursing care intensity for different patient groups is a
large factor in the needs of staff resources. This does not
mean how demanding the medical treatment is, since a
patients need for care may be more demanding than what a
medical diagnosis might predict (Christ-Grundmann 1997,
Van Slyck & Johnson 2001). In recent years, patient
classification systems have been developed in a new way
due to improved information systems that facilitate the
receiving of complete information relating to patients, wards
and hospitals (Levenstam & Bergbom Engberg 1997,
Bjorkgren et al. 1999, Mueller 2000). It is possible to analyse
and use these comprehensive information sources in the
planning and production of health care, as well as evaluating
its productivity (Hofdijk 1997, Sanderson & Mountney
1997).
Classification systems of health care have improved
throughout recent decades. They are usually based on
medical diagnosis, such as ICD-10. Diagnostic-RelatedGroups (DRG), a system for describing hospital production
developed in the USA, is a good example of such a system
combining clinical diagnosis and treatment. It determines the
amount of resources used to care for a patient within a
certain ICD-10 category (Fetter & Freeman 1986, Freeman
et al. 1995, Muldoon 1999). The DRG cost weights are
country specific and based on statistical analysis of the
costing data in each country (Mikkola et al. 1998). Implementation of DRG in the 1980s strongly influenced the
development of PCS because measurements data based on
medical diagnosis do not show the information for nursing
management and its productivity and costs and do not solve

the problem of allocating nursing resources optimally to


ensure qualitative care. (Malloch & Conovaloff 1999, Averill
& Fairbrother 2000)
The only way to measure the work done by nurses from a
patient perspective and on a hospital or ward level is to use a
patient classification system where the information on the use
of resources can be used in staff management and cost
accounting (Thomas & Vaughan 1986, Hoffman 1988). In
this study, we use a patient classification system to analyse
the use of nursing resources. A patient classification system
(PCS) assesses and classifies patients according to their need
of care as well as the activities that are necessary to fulfill the
needs of the care process during a certain time period. It can
be determined by methods and processes that are used to
identify, validate and monitor the needs of an individual
patient and provide information for human resource administration, accounting, budgeting and other functions of
management. (Huhkabay & Skonieczny 1981, De Groot
1989a,b)
The aim of this study is, through one case, to assess how
health-care managers have used the new Finnish PCS in
nursing care management during a two-year period. Active
management should result in the transferring of nursing
resources from wards with a low nursing care intensity per
nurse ratio (situation below the optimal nursing level) to
wards with a high nursing care intensity per nurse ratio
(situation above the optimal nursing level). The study aims
can be summarized in the following study questions:
1 How feasible is the used PCS in nursing management to
measure the utilization of resources and their cost implications?
2 Does the PCS reflect the differences in utilization of nursing
resources between different general inpatient wards?
3 How much resources have been saved or lost in nursing
care management when the nurses working days are too
busy or too quiet, in other words when the nursing care
intensity per nurse ratio has been above or below the
optimal level in each ward?
4 Has there been active nursing management through the
transfer of nursing resources in two consecutive years
depending on the need for patient care, in other words
from wards below the optimal level of nursing
care intensity per nurse to wards with above optimal
levels?

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

675

A-K Rainio and AE Ohinmaa

Methods
Presentation of the PCS (RAFAELA-system)
The development of the patient classification system (PCS)
at Vasa Central Hospital started in 1992 and it was taken
into experimental use in all hospital patient wards in
19941995. The PCS received the name RAFAELA from
the research team who developed it: Rainio Fagerstrom
Rauhala. (Rainio 1999) The RAFAELA-system consists of
three parts: Oulu Patient Classification (OPC) system
(Kaustinen 1995), nurse resource registry, and Professional
Assessment of Optimal Nursing Care Intensity level
(PAONCIL) measure. Using the OPC system and nurse
resource registry, we are able to calculate nursing care
intensity points per nurse during each day. The PAONCIL
measure is used to estimate, using linear regression analysis, the optimal nursing care intensity level per nurse,
which describes the needed nurse resources in proportion
to the need of patient care. (Fagerstrom & Rainio 1999a,
Fagerstrom et al. 2000b).

The RAFAELA PCS can be described by the following


formula:
RAFAELA PCS (OPC/Nursing resource) and PAONCIL
method.
The RAFAELA nursing care intensity classification system
consists of data shown in Fig. 1. The data are part of the
hospitals electronic patient administration system. The
RAFAELA system is not described in detail but previous
articles are referred to.

Oulu Patient Classification (OPC)


The OPC (Oulun yliopistollinen keskussairaala 1994, Kaustinen 1995) consists of six subsections of nursing care:
(1) planning and co-ordination of care; (2) breathing, blood
circulation and symptoms of disease; (3) nutrition and
medication; (4) personal hygiene and excretion; (5) activity,
movement, sleep and rest; (6) teaching, guidance in care and
follow-up care, emotional support. In each of these areas the
nurse classifies patients into one of four classes according to
their need of care (from A 1 point to D 4 points). In the

HOSPITALS ELECTRONIC RECORDS


PATIENT MANAGEMENT
Inout patient
Resource booking
Waiting lists
Patient discharge and invoice

OPC
(Patient classification)

ADMINISTRATION
Accounting
Human resources

RESOURCES

Nursing care intensity scores


depending on patients need
for care / ward / day

Nurse resources / day


Salary + benefits

NURSING CARE MANAGEMENT REPORT


FROM RAFAELA PCS
Nursing care intensity
scores per
patient
nurse
day / month / year
patient room
DRG group

HILMO
Discharged
patient
record
(national
registry)

Nursing care intensity per


nurse for personal administration / day / month / year
days below the optimal
level
days at the optimal level
days above the optimal
level

PAONCIL

(Professional Assessment of Optimal Nursing Care Intensity


Level)

-8 weeks
Research in the wards during 68
weeks

Figure 1 RAFAELA PCS as a part of the


electronic administrative system in Vasa
Central Hospital.
676

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

Issues in clinical nursing

classification a 10-paged classification manual is used. The


OPC summary score is calculated by adding up the points in
the six care areas of the system. Thus the care intensity of a
patient can be 624. The higher the score, the higher the
nursing care intensity of the patient. The OPC classification is
based on the principles of nursing presented in the quality
control programme and on Ropers model of nursing. (Oulun
yliopistollinen keskussairaala 1994) The development of the
classification system is based on the Hospital Systems Study
Group (HSSG) classification (Goldstone et al. 1983).
The OPC measure is used once a day. The OPC measure
does not describe how nurses have filled their day with
different tasks, but how they have answered the patients needs
with the different processes and interventions of nursing work
(Hoffman 1988, Rainio 1994, Kaustinen 1995, Rainio 1998).
The validity and reliability of the OPC measure has been tested
in previous studies (Fagerstrom et al. 2000a, Fagerstrom &
Rauhala 2001, 2003). Even from the point of view of the
patient and the nursing care intensity content of the measure
has been improved (Fagerstrom et al. 1998, 1999b).

Nursing resources
Each day the number of nurses who have worked with the
patients in a ward is registered into the nurse resource registry
(Fig. 1). The OPC score is divided by the number of the
nurses in the ward on each day. The nursing care intensity
point per nurse ratio describes the productivity of nursing
care in the ward (Rainio 1999, Fagerstrom & Rainio 1999a,
Fagerstrom et al. 2000b).

Professional Assessment of Nursing Care Intensity Level


(PAONCIL)
The third part of the RAFAELA PCS is PAONCIL and it
answers the question, when does the nursing care intensity/
nurse ratio reach the level of good care, in other words when
do nurses estimate that they have given their patients good
care? PAONCIL is an alternative to observation research and
time measurement. Nurses assess how they have experienced
their shifts workload using a seven-class measure. Zero is the
optimal situation, 13 is a situation where nurses have too
much work and the bigger the figure is, the more nurses have
to prioritise tasks. Minus three to 1 describes a situation
where the closer nurses are to 3, the more they feel they
have time for less important tasks. Classification guidelines
have been developed for the whole numbers of the scale.
Material is gathered during a period of approximately two
months, so that it would include an estimate of several
hundred shifts depending on the ward size and number of

Assessment of nursing management and utilization of nursing resources

nurses. Using a linear regression analysis it is possible to


examine to what extent the nursing care intensity/nurse ratio
explains the workload experienced by the nurses (PAONCIL
score). The regression model is then used to estimate the level
of nursing care intensity per nurse ratio where the corresponding PAONCIL value is zero. With an administrative
decision the studys results have been used so that when the
nursing care intensity point per nurse is, e.g., 27, the optimal
nursing care intensity level in the ward can vary 3.5 points,
indicating that the optimal range would be between 23.5 and
30.5 nursing care intensity point/nurse per day. There will
always be a variation between days in the nursing care
intensity point; which variation is acceptable is decided
democratically by nurse managers. The decision is based on
practical management, no scientific method has been found
by researchers to do that. This nursing work assessment
method was developed in Vasa Central Hospital (Rainio
1996, 1999, Fagerstrom & Rainio 1999a, Fagerstrom et al.
2000b). The validity of the content of the PAONCIL method
has been tested by an expert group of 150 nurses. The
classification has been found universally applicable and its
validity acceptable. (Fagerstrom et al. 2002).
In the RAFAELA system each ward has its own optimal
level of nursing care, which is estimated using the PAONCILmeasure (Table 1). It indicates a situation where nurses assess
that they have given the patients good quality care.

RAFAELA and nursing benchmarking in Finland


The aim of the RAFAELA-system has been, with respect to
nursing philosophy and for the purpose of using the data, to
give a picture of modern, professional nursing care. The longterm development of the system has been based on the
philosophy of good, comprehensive care for the patient.
The RAFAELA-system has been tested nationally using
material from seven hospitals during the years 20002001.
The study has researched the state of Finnish nursing and
benchmarking data using the RAFAELA-system. In the study,
most specialized fields of nursing had a nursing care intensity
ratio above the optimal area during approximately 4050%
of days, and below the lowest limit during 1020% of days.
There were differences between specialized fields and, for
example, the nursing care intensity in the paediatric wards
was above the optimal area during 25% of days and below
the optimal area during 50% of days (Fagerstrom & Rauhala
2001, 2003). From the point of view of nursing management,
data from many hospitals do not show the efficiency or
activity of the management of an individual hospital, but
show the special health care situation on a national level and
offer unique reference material.

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

677

A-K Rainio and AE Ohinmaa


Table 1 Beds, numbers of patients, inpatient days, average number of nurses/day and average nursing care intensity/nurse in the wards and the
optimal level in nursing care intensity by PAONCIL-method
Admitted
patients

Inpatient days

Number of
nurses/day

Nursing care
ntensity point/nurse

Ward

Beds
inward

2000

2001

2000

2001

2000

2001

2000*

2001*

Optimal
level**

A
B
C
D
E
F
G
H
I
J
K
L

20
27
28
27
15
15
30
32
20
24
14
30

1026
2247
2337
1840
848
839
1741
1725
1287
1786
1329
2105

1185
2406
2823
1931
742
1070
1771
1719
1372
1745
1497
2508

8606
10900
12049
11476
4782
4334
10269
11925
7033
7399
5221
8994

7516
10448
11646
11268
4111
5287
10497
11763
7326
7233
5629
9898

9.59
10.6
11.1
12
7.21
7.09
12.3
14.4
11.6
10.5
7.84
20.5

9.66
10.8
11.5
12.9
6.97
7.49
12.5
14.7
11.9
10.7
8.04
20.9

26.11
30.66
27.79
31.36
28.28
26.95
28
28.28
21.35
29.19
19.53
14.98

27.58
31.71
29.75
32.76
27.51
27.72
27.65
28.21
22.54
29.4
20.72
15.19

20,327,3
20,327,3
2128
22,429,4
23,130,1
23,130,1
21,728,7
22,429,4
17,524,5
26,633,6
18,925,9
11,218,9

*An average value on each year.


**Determined in a consensus by nurse administrators.

Data
The study data have been collected from Vasa Central Hospital
in Finland, which includes 24 specialized care wards. The
hospitals catchment area comprises approximately 170 000
people. The data consist of information collected from 12
general specialized health care wards from the years 2000 and
2001, excluding psychiatric specialized care. The researched
wards are four internal medicine wards (AD), two oncologic
wards (E and F), four surgical wards (GJ), a prenatal and
gynecological ward (K) and a paediatric ward (L).
The line of nursing management is formed in the following
way: the head nurse of a ward is responsible for the nursing
of an entire ward, the nurse manager is responsible for several
care units and the director of nursing is responsible for
nursing in the entire hospital and is on the executive board.
Table 1 shows the researched wards numbers of patients,
bed days, average numbers of nurses per day and average
nursing care intensity/nurse ratio (OPC/nursing resources) for
2000 and 2001 as well as the level of optimal nursing care
intensity per nurse (PAONCIL) in the different wards. The
number of nurses has changed most in ward D, where in
2001 there was on average one more nurse/day than in 2000.

Cost estimations in RAFAELA


The nursing management of inpatient wards was analysed in
the context of used resources and RAFAELA PCS, and the
one-day wage cost (W) in each ward was calculated using the
following formula: W (XZ)/366, where:
678

X is the wards gross salaries during one year including


benefits;
Z is the employers social insurance reimbursement of
nurses sickness leaves;
366 is the number of days during year 2000 (365 in 2001).
In the analysis of the total cost of care (Y) in each ward,
only the days that were included into the RAFAELA data (D)
were included (Y W * D).
The RAFAELA system was used to calculate the nursing
care intensity points (T) of the days registered in the system
(D). The cost of one nursing care intensity point (S) was
calculated by dividing the total cost of care Y by nursing care
intensity points T during a year (S Y/T).
The cost of one nursing care intensity point (S) describes
how much resources (in monetary units) have been used to
satisfy the patients need for care. It is assumed that one
nursing care intensity point is equivalent in each ward, that is,
the way the assessment reflects the response of a nurse to the
needs of a patient is similar between the patients in the wards.
The analysis of the nurse management in the context of costs
and the nursing care intensity per nurse ratio was carried out
by comparing the actual nursing care intensity level on one
day (from PAONCIL) to the optimal range in the ward
(Table 1).
First the data were analysed to find out what resources the
employer had lost financially and in nurse numbers during the
days when the nursing care intensity/nurse ratio had been
below the optimal level. For example, the minimum of the
optimal level in ward E is 23.1 and a certain days measured
ratio is 17.7 (a difference of 5.4). When this figure is

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

Issues in clinical nursing

Assessment of nursing management and utilization of nursing resources

multiplied by the number of nurses that day, we get the


nursing care intensity point which is needed to reach the
minimum of the optimal level (23.1). When this nursing care
intensity point is multiplied by the price of ward Es nursing
care intensity point (6,35 in 2000 and 6,65 in 2001), we
reach the amount that the employer had lost with inefficiently
used nurse resources.
In the analysis we added up the days below the optimal
level in each ward. This figure was then divided by a nurses
average basic labour costs without benefits. This shows the
employers loss in nurse resources. Secondly, we analysed the
days that had exceeded the optimal level, that is the days
when the employer had saved on nurse resources, in the same
way.
For example, the upper limit for the optimal limit in ward
E is 30.1, and the nursing care intensity/nurse ratio for a
certain day is 31 (a difference of 0.9). When this figure is
multiplied by the number of nurses that day, we obtain the
figure which is needed to reach the optimal level (30.1) as

nursing care intensity points. When this figure is multiplied


by the cost of the wards nursing care intensity point (6,35 in
2000 and 6,65 in 2001), we find out the employers financial
saving. When we add up the savings of the employer during a
year and divide it by the basic salary of a nurse, we get an
estimate of the number of additional nurses needed in the
ward to reach the optimal nursing care intensity per nurse
level.

Results
The average numbers of nursing care intensity points per
nurse were on the optimal level on 10 out of 12 wards in
2000 and eight wards in 2001 (Table 1). In nine wards the
average nursing care intensity level/year has risen from 2000
to 2001.
Table 2 shows that the labour costs of nursing have risen
from 2000 to 2001 in all wards except E, K and L. Income
from medical insurance has risen in 2001. The number of

Table 2 Labour costs (X), health insurance income (Z), daily labour costs (W), days registered in the RAFAELA (D) and their costs (Y), nursing
care intensity points during a year in the ward (T) and the cost of one nursing care intensity point (S) during 2000 and 2001
Z Sickness
leave

W (N 366)
Salary/day

Measured
days (D)

Cost of
excluded days

Y Measured
cost

T Care intensity
points

S (Y/T) Cost
of point

Year 2000 wards


A
690755
B
723690
C
755628
D
655354
E
462160
F
419036
G
771721
H
937118
I
806194
J
564592
K
578883
L
1541125
Total
8906257

5458
8737
11827
2084
8440
2705
6050
21639
3016
10439
9598
25004
114997

1872
1953
2032
1785
1240
1138
2092
2501
2194
1514
1555
4142
24020

162
364
365
273
341
321
362
362
366
268
366
364
3914

381969
3907
2032
3570
30992
51188
8368
10005
0
148380
0
8285
648695

303328
711046
741768
649700
422728
365142
757304
905475
803178
405773
569285
1507837
8142564

38580
114409
109544
102501
66564
58022
116962
140253
45928
80364
54435
105173
1032735

7.86
6.21
6.77
6.34
6.35
6.29
6.47
6.46
17.49
5.05
10.46
14.34
7.88

Year 2001 wards


A
697160
B
760120
C
838180
D
846915
E
448936
F
473899
G
825016
H
958255
I
854834
J
578929
K
572321
L
1513147
Total
9367713

14051
11277
8030
8321
3408
14660
14592
14334
10370
3454
5656
12821
120975

1872
2052
2274
2298
1221
1258
2220
2586
2314
1577
1553
4110
25264

364
361
365
365
297
364
364
365
365
275
355
363
4203

1872
8207
0
0
83002
1258
2220
0
0
141898
15525
8221
262203

681238
740636
830149
838594
362525
457981
808204
943920
844464
433578
551140
1492105
8984534

91920
117937
121310
146163
54539
71631
121206
142809
94397
81154
57924
109148
1210138

7.41
6.28
6.84
5.74
6.65
6.39
6.67
6.61
8.95
5.34
9.51
13.67
7.42

X Salaries

The hospitals expenses in 2000 were 90 897 434 and 96 774 312 in 2001.
 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

679

A-K Rainio and AE Ohinmaa

days registered in the data rose in 2001. In 2001 the cost of


care intensity points went down in wards A, D, I, K and L, in
other words in proportion to costs the patients needs were
answered at a lower cost in 2001 compared to 2000, even
though Table 1 shows a slight increase in average nurse
resources. The cost of care intensity points had increased in
seven wards, in other words the patients needs were fulfilled
at a higher cost than in 2000. The change in the cost of care
intensity points does not seem to have a connection to the
changes in average nurse resources shown in Table 1. The
changes in the cost per nursing care point within the wards
were small except in surgical ward I (8.54 points), which had
an exceptionally small number of surgeries during 2000 and
can be considered as an exception in the data. If this ward is
excluded from the average cost of a nursing care point, there
was a 0,13 decrease in the average nursing care point cost.

Table 3 shows that wards operated for approximately


77 days (21%) and 71 days (19%) on average below the
optimal nursing care intensity range in 2000 and 2001,
respectively. The period above the optimal level increased
from 106 days (29%) to 129 days (35%) on average. When
these days were transformed into the number of nurses per
year in resource utilization below or above the optimal
resource, the employer can be seen to have lost a resource of
9.84 nurses in 2000. Transformed to monetary units this is
equivalent to 307,745. Respectively in 2000, there was a
shortage of 11.80 nurses (369,080) in relation to the need of
patient care. More effective allocation of resources would
have been needed, especially in wards B, C, D, G and H,
where the employer had saved more in resources than what
had been lost, and in wards JL, where resources were
inefficiently utilized. The resources needed for patient care

Table 3 Nurses operating below and above the optimal nursing care intensity range measured by days, wages and number of nurses in days
registered to RAFAELA during 2000 and 2001
Nurses operating below optimal level

Nurses operating above optimal level

Days

Lost wages

Nurses/year

Days

Saved wages

Nurses/year

Year 2000 ward


A
162
B
364
C
365
D
273
E
341
F
321
G
362
H
362
I
366
J
268
K
366
L
364
Total
3914
Average

26
26
31
15
103
123
61
56
86
125
187
90
929
77

4096
6156
5077
5324
20291
29261
17388
11294
38285
44627
66497
59450
307745
25645

0.13
0.20
0.16
0.17
0.65
0.94
0.56
0.36
1.22
1.43
2.13
1.90
9.84
0.82

43
249
144
191
111
86
128
109
73
45
44
47
1270
106

9209
86923
31750
53329
25122
18982
37822
25576
35149
7450
9683
28084
369080
30757

0.29
2.78
1.01
1.70
0.80
0.61
1.21
0.82
1.12
0.24
0.31
0.90
11.80
0.98

Year 2001 ward


A
364
B
361
C
365
D
365
E
297
F
364
G
364
H
365
I
365
J
275
K
355
L
363
Total
4203
Average

50
15
11
11
104
131
60
63
66
123
146
70
850
71

11171
3466
1684
1783
22800
32564
13328
26078
17153
36545
36049
39524
242143
20179

0.35
0.11
0.05
0.06
0.71
1.02
0.42
0.82
0.54
1.14
1.13
1.24
7.58
0.63

147
259
220
254
88
119
124
117
96
52
44
33
1553
129

39789
96538
54503
85672
19336
27050
32594
30647
28523
11512
9651
21799
457615
38135

1.25
3.02
1.71
2.68
0.61
0.85
1.02
0.96
0.89
0.36
0.30
0.68
14.32
1.19

RAFAELA
Days

The labour costs of one nurses basic salary in 2000 was 31 283 and in 2001 31 956.
680

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

Issues in clinical nursing

would have partly been in use in general wards if they had


been more effectively used by operational management. This
would have required transferring nurses to new wards and
the hiring of two new nurses.
In 2001 the employer lost 7.58 nurses in resources. This is
equivalent to a cost of 242,143 for these days below the
optimal level. There was a shortage of 14.32 nurses
(457,615) in relation to the need for patient care. The
working pace seems to have increased especially in all the
wards for internal medicine and surgical ward G. In 2001
the need to hire new nurses in the hospital increased
compared to the previous year.
The data do not directly show the active operative
management of nursing or the transferring of nursing
resources according to the patients need for care over time.
From 2000 to 2001 ward D received the additional work
contribution of 0.9 nurses/day on average. In other wards the
changes in nurse resources were below 0.5 nurses/day on
average (Table 1). In wards AD the situation became more
difficult in 2001 compared to 2000. In wards EI the
situation remained approximately the same and differences
in nurses operating above and below the optimal nursing care
intensity range was small. In wards JL nurses operated
below the optimal range more frequently than above it,
indicating that these wards had some excess nursing
resources.

Discussion
The essential purpose of the RAFAELA-system is to use
information in the allocation of nursing resources to where
the good quality of patient care requires it, in other words to
work as an instrument for effective nursing care management. The study shows that the RAFAELA-system is suitable
for analyzing nursing management and its cost effects.
The wards of the researched hospital did not completely
succeed in the allocation of nursing resources according to
the need of patient care. Also, the workload in nursing
increased from 2000 to 2001, as nurses operated above the
optimal limit more than below it. There were not always
nurses where they would have been needed, in other words
there have been problems in ensuring the quality of the care
of patients. On the other hand, the employer lost nursing
resources in almost all the wards, because during both years
the level was below the optimal limit and the nurses had
inefficient working time.
The results are inconclusive, because data from a period of
only two years were used and the data cover only the general
wards of one central hospital. However, this case study
shows how the PCS could be used to assess active nursing

Assessment of nursing management and utilization of nursing resources

care management. The optimal upper and lower limits of the


PAONCIL-measure have been set by an administrative
decision, so it is impossible here to assess how correct they
are. The principle was to treat each ward according to the
same criteria (Rainio 1999). The RAFAELA-system has been
shown to possess a high level of validity, (Fagerstrom et al.
1998, Fagerstrom et al. 2000b) and its development and
testing has continued in a national program (Fagerstrom &
Rauhala 2001, 2003). The recent result of the national data
shows that the average reliability of parallel classification was
77%, which can be considered a good result. However, the
validity of hospitals in the national reference material was
weakened by the variation of coding practices in different
hospitals (Fagerstrom & Rauhala 2003). The reliability and
usability of the RAFAELA-system has been tested on the
material of eight Finnish hospitals and it has been found a
reliable tool in personnel planning, budgeting and cost
calculation (Rauhala & Fagerstrom 2004).
A lack of nursing resources in the short term existed in
almost all the wards at some point. The question is, in
nursing, is there a daily, operative management, with which
nursing resources can be allocated according to the need for
patient care? In answering this question there is the problem
that nursing care intensity classification systems show the
situation in retrospect and are not able to predict the
situation of the future (Hoffman 1988). In addition, we must
acknowledge other problems of staff management in acute
specialized health care. Three-shift-work sheets planned in
advance and quickly changing situations make successful
operative management difficult to achieve (Mark 2002).
However, these kinds of data from several years offer
material needed for strategic staff management. Data
received from nursing care intensity classification systems
should be used diversely in the decision-making of nursing
management (Van Slyck 2000). Nursing resources that have
been used are registered, but this registry could also be used
to show possible transfer of nurses, temporary or permanent,
to other wards. This would show more clearly the daily
management decisions in nursing care.
The limited scope of the nursing care intensity classification
system must be admitted. It shows the patients acute care
needs but not the many other factors which affect the needs
of nursing resources in nursing units (Stepura & Miller
1989). Other factors that affect nursing (such as the planning
of work, shift rotas, substitutes, students, co-operation
between professions, ones own capacity for work, stress)
have been taken into consideration in the PAONCIL method
and thus an attempt to separate the meeting of patients needs
and other factors adding to the work load has been made
(Rainio 1999, Fagerstrom et al. 2002). The PCS is not a

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

681

A-K Rainio and AE Ohinmaa

measure of a wards nursing care intensity, but of the patient


(Van Slyck 2000). In the management of nursing, diverse
information sources of the wards situation and productivity
should be used (Barton 1994) such as treatment periods, bed
days, out-patient visits, DRG-groups and national benchmarking information.
The data raise the question as to whether the management
of nursing work is sufficiently interested in real information,
rather guided more by subjective, experience based information than objective, measurable and verified information, as
Vuori (2001) has suggested. The rigidity of the wards in
adding nursing resources during the second year, even though
data from the first year indicated a clear lack of nurses, shows
that not all the levels of management in the hospital acted
according to the results. Annual and monthly reports on the
situation in nursing should be made. These would show
which way nursing in wards is developing, whether nursing
resources are in optimal use and whether the work is
significantly under or over resourced. This information
should also be used as a part of the hospitals own budget
planning, because significant changes in nursing resources
mean significant changes in the wards costs.
When we look at the costs of nursing, we can say that the
work pace of nurses had, looking at total costs, increased. In
2001 the employer saved more than in 2000. In 2000 the
savings and losses were fairly close to each other, but in 2001
the losses were only half of the savings. Roughly estimated on
the basis of the patients nursing care intensity, every ward
would have needed one more nurse to ensure the quality of
care. The nurses also had ineffective working time, which is a
sign that adding to resources does not solve the managerial
problem of the right allocation of resources. If resources are
placed wrongly, the problems of daily staff management
continue.
Vuori (2001) states that the adding of resources does not
guarantee that they are used appropriately. Management
should learn to interpret information and assessment data in
order to make management more effective. The problems of
health care are not solved by changing the administration or
by adding to resources, but by questioning the things that are
taken for granted in the operational models of organizations
and work communities. Administration needs both clinical
and managerial leadership, so that hospitals can reach
successful solutions (Ham 2003).
In the researched wards where the nursing care intensity
per nurse ratio was constantly high, we can predict that the
nurses will soon be exhausted, take sick leave or move to
wards with a lower nursing care intensity level. The nurses
evaluating the PAONCIL method stated that calm periods
(days below the optimal level) are necessary for relaxation
682

and re-balancing after busy days (Fagerstrom et al. 2002). It


is likely that nursing intensity and necessary nursing resources
are connected to nursing satisfaction and turnover (Mark
2002). The greatest motivation for hospital staff is giving
high quality services to patients (Ham 2003). High nursing
care intensity (days above the optimal level) was described by
the nurses as a feeling of inadequacy when good quality of
care cannot be quaranteed (Fagerstrom et al. 2002). It is not a
good sign for the quality of nursing that nursing resources are
for several years used to the limit. Especially in wards BD
the significant need for additional resources during 2000 did
not lead to new recruitments except in ward D, where there
was an average addition of 0.9 nurses/day, and the above
optimal level time still increased during the next year. In
addition to the likely decreased quality of care, it is possible
that this has also had an impact on the well-being of nurses
over time. Sick leave reimbursements have increased a small
amount from 2000 to 2001 and there seem to be more
increases in reimbursements amongst wards that have a
higher need for additional nursing resources. However, this
pattern is not true in every ward.

Conclusions
The first goal of this study was to show how feasible the
RAFAELA PCS is in nurse management. On the basis of this
case, it seems that RAFAELA can be used to measure the
nursing care intensity/nurse ratio in different wards and
further receive annual reference data needed in the management of nursing.
The second goal was to test whether RAFAELA reflects the
differences between wards. The analysis shows that there are
significant differences between wards that remained almost
unchanged during the two-year period.
The third goal was to analyse how much the employer has
saved or lost in nursing costs and resources. The study shows
that the employer saved more in nursing resources than what
was lost. There were significant differences between wards.
The fourth goal was to research whether active nursing
management could be found in the data. The different
nursing care intensity per nurse levels in different wards did
not improve during the two years, nor did the situation in the
higher care intensity level wards. It is noteworthy that the
general situation of nursing seemed to become more difficult
in 2001 compared to 2000.
According to this research, data from PCS should be used
in staff management, otherwise its implementation in the
organization is not justifiable. The nursing resources could be
better allocated by the management according to patient
needs. Nurses have either too heavy or too light days, and the

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

Issues in clinical nursing

quality of care decreases during busy days, when nursing


resources lost from the other wards with light days would
have been needed.
RAFAELA PCS and the methods and registers it is based on
should be further researched and improved to ensure reliability and validity. Decision-making concerning the whole
hospital requires reliable information about the unit where
additional resources need to be allocated. In health care the
success of management in cost control, productivity and the
effective allocation of staff is evaluated using more and more
sources of information. In the RAFAELA system the productivity of nursing can be seen as nursing care intensity per
nurse figures and can be directly converted into money and
numbers of nurses. This information supports nursing managers in the fair allocation of resources and in the guaranteeing of the quality of care.

Acknowledgements
Sincere thanks to Vasa Central Hospital for the opportunity
to use data from the RAFAELA PCS. Thanks also to Olle
Pursiainen, the IT-manager, for helping with Fig. 1. From
2003 the RAFAELA PCS has been owned by The Association
of Finnish Local and Regional Authorities. The study was
supported by the Medical Research Fund of Vasa Hospital
District.

Contributions
Data collection: PP, AR; data analysis: AR, UI; manuscript
preparation: ALM; translation: JS.

References
Averill C & Fairbrother M (2000) Developing a statewide patient
classification system. Nursing Administration Quartely 24, 29
35.
Barton A (1994) Data needs for decision support of chief nurse
executives. JONA 24 (suppl. 4), 1925.
Bjorkgren MA, Hakkinen U, Finnet-Soveri H & Fries BE (1999)
Validity and reliability of Resource Utilization Groups (RUG-III) in
Finnish long-term care facilities. Scandinavian Journal of Public
Health 27, 228234.
Christ-Grundmann D (1997) Patient acuity: a viable methodology for
linking patient and financial data. Aspenss Advisor for Nurse
Executives, August, 46.
De Groot HA (1989a) Patient classification system evaluation. Part 1:
essential system elements. JONA 19, 3035.
De Groot HA (1989b). Patient classification system evaluation. Part
2: system selection and implementation. JONA 19, 2430.
Fagerstrom L, Bergbom Engberg I & Eriksson K (1998) A comparison between patients experience of how their caring need have

Assessment of nursing management and utilization of nursing resources


been met and the nurses patient classificationan explorative
study. Journal of Nursing Management 6, 369377.
Fagerstrom L, Bergbom Engberg I & Eriksson K (1999b) The
patients perceived caring needs: measuring the unmeasurable.
International Journal of Nursing Practice 5, 199208.
Fagerstrom L & Rainio A-K (1999a) Professional assessment of optimal nursing care intensity level: a new method of assessing personnel
resources for nursing care. Journal of Clinical Nursing 8, 369379.
Fagerstrom L, Rainio A-K, Rauhala A & Nojonen K (2000a) Validation of a new method for patient classification, the Oulu Patient
Classification. Journal of Advanced Nursing 31, 481490.
Fagerstrom L, Rainio A-K, Rauhala A & Nojonen K (2000b)
Professional assessment of optimal nursing care intensity level.
Scandinavian Journal of Caring Sciences 14, 97104.
Fagerstrom L & Rauhala A (2001) Finnhoitoisuus-hoitotyon
benchmarking. Pilottiprojektin raportti vuodelta 2000. Suomen
Kuntaliitto, Helsinki. (in Finnish)
Fagerstrom L, Nojonen K & Akers A (2002) Metodologinen triangulaatio PAONCIL-menetelman sisallon validiteetin testaamisessa.
Hoitotiede 14, 180191. (in Finnish)
Fagerstrom L & Rauhala A (2003) Finnhoitoisuus-hoitotyon
benchmarking. Projektin loppuraportti vuodelta 20002002.
Suomen Kuntaliitto, Helsinki. (in Finnish)
Fetter R & Freeman J (1986) Diagnosis related groups: product line
management with hospitals. Academy of Management Review 11,
4154.
Freeman J, Fetter R, Park H, Schneider K, Lichtenstein J, Hughes J,
Bauman W, Duncan C, Freeman D & Palmer G (1995) Diagnosisrelated group refinement with diagnosis- and procedure-specific
comorbidities and complications. Medical Care 33, 806827.
Fare R, Grosskopf S, Lindgren B & Poullier J (1997) Productivity
growth in health-care delivery. Medical Care 35, 354366.
Goldstone LA, Ball JA & Collier MM (1983) Monitor, an Index of
the Quality of Nursing Care of Acute Medical and Surgical Wards.
Polytechnic Products LTD, Newcastle Upon Tyne.
Ham C (2003) Improving the performance of health services: the role
of clinical leadership. The Lancet 361, 19781980.
Hofdijk J (1997) The development of patient groupings for more
effective management of health care. A comment. European
Journal of Public Health 7, 210214.
Hoffman F (1988) Nursing Productivity Assessment and Costing Out
Nursing Services. J.B. Lippincott company, Philadelphia.
Huhkabay LM & Skonieczny R (1981) Patient classification systems:
the problems faced. Nursing and Health Care February, 89103
Kaustinen T (1995) Hoitoisuusluokituksen kehittaminen ja arviointi.
Lisensiaattitutkielma. Hoitotieteen laitos, Oulun yliopisto. Oulu,
Finland. (in Finnish)
Levenstam A-K & Bergbom Engberg I (1997) How to translate
nursing care into costs and staffing requirements: part two in the
Zebra system. Journal of Nursing Management 5, 105114.
Linna M (1998) Measuring Hospital Performance: The Productivity,
Efficiency and Costs of Teaching and Research in Finnish Hospital.
National Research and Development Centre for Welfare and
Health. Research report 98. Helsinki, Finland.
Linna M & Hakkinen U (1999) Determinants of cost efficiency of
Finnish hospitals: A comparison of DEA and SFA. Helsinki
University of Technology, Systems Analysis Laboratory, Research
Report A78.

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684

683

A-K Rainio and AE Ohinmaa


Malloch K & Conovaloff A (1999) Patient classification systems,
part 1: the third generation. JONA 29, 4956.
Mark BA (2002) What explains Nurses perceptions of staffing
adequacy? JONA 32, 234242.
Mikkola H, Keskimaki I & Hakkinen U (1998) Facts about DRGs
From Experiences Elsewhere to Applications in Finland. Aiheita
39, Stakes, Helsinki, Finland.
Mueller C (2000) The RUG-III case mix classification system for
long-term care nursing facilities. JONA 30, 535543.
Muldoon JH (1999) Structure and performance of different DRG
classification systems for neonatal medicine. Pediatrics 103, 302
318.
Oulun yliopistollinen keskussairaala (1994) Hoitoisuusluokituksen
kehittaminen osaksi potilasryhmakohtaista kustannuslaskentaa
Oulun yliopistollisessa keskussairaalassa. Tutkimus- ja kehittamishankkeen loppuraportti. Oulu, Finland. (in Finnish)
Rainio A-K (1994) Hoitoisuusluokituksen eli potilaiden yksilollisen
hoidon tarpeen mittaamisen kehittaminen. Hoitotyon projekti,
hoitoisuusmitoituksen kehittaminen, raportti II, Vaasan sairaanhoitopiiri. Vaasa, Finland. (in Finnish)
Rainio A-K (1996) Hoitoisuusluokituksen hyodyntaminen erikoissairaanhoidon kustannuslaskennassa. Hoitotiede 8, 7077.
(in Finnish)
Rainio A-K (1998) Hoitoisuusluokituksen kayttoonotto. Kokemuksia
Vaasan keskussairaalasta. Sairaanhoitaja 71, 1820. (in Finnish)

684

Rainio A-K (1999) Luotettavan hoitoisuusluokitusjarjestelman


implementointi Vaasan keskussairaalassa. Vaasan sairaanhoitopiirin julkaisuja 2. Vaasa, Finland. (in Finnish)
Rauhala A & Fagerstrom L (2004) Determing optimal nursing
intensity: the RAFAELA method. Journal of Advanced Nursing 45,
351359.
Sanderson H & Mountney L (1997) The development of patient
groupings for more effective management of health care. European
Journal of Public Health 7, 210214.
Stepura B & Miller K (1989) Converting nursing care cost to revenue.
JONA 19, 1822.
Thomas S & Vaughan R (1986) Costing nursing services using RVUs.
JONA 16, 1016.
Van Slyck A (2000) Patient classification systems: not a proxy for
nurse busyness. Nursing Admininistration Quartely 24, 6065.
Van Slyck A & Johnson K (2001) Using patient acuity data to
manage patient are outcomes and patient care costs. Outcomes
Management for Nursing Practice 5, 3640.
Vuori J (2001) Oppimisen ja johtamisen mahdottomuudesta terveydenhuollon organisaatioissa. Terveyshallinnon popperilainen
falsifikaatioteoria osa 1. Learning and leadership in health care
organisation. Poppers theory of falsification applied in health
management. Hallinnon tutkimus 2, 166187. (in Finnish)

 2005 Blackwell Publishing Ltd, Journal of Clinical Nursing, 14, 674684