Journal of Evaluation in Clinical Practice ISSN 1356-1294

Evaluation of documented nursing care plans by the use of nursing-sensitive outcome indicators
jep_1233 611..618

Inger Jansson RN,1 Ewa Pilhammar-Andersson PhD2 and Anna Forsberg PhD3
1 Doctoral Student, School of Social and Health Sciences, Halmstad University, Halmstad and Institute of Health and Care Sciences, The Sahlgrenska Academy at Gothenburg University, Gothenburg, Sweden 2 Professor, Institute of Health and Care Sciences, The Sahlgrenska Academy at Gothenburg University, Gothenburg, Sweden 3 Senior Lecturer, Institute of Health and Care Sciences, The Sahlgrenska Academy at Gothenburg University, Gothenburg, Sweden

Keywords care plans, empirical research report, nursing diagnosis, outcome, stroke Correspondence Inger Jansson School of Social and Health Sciences HOS Halmstad University Box 823 SE-301 18 Halmstad Sweden E-mail: inger.jansson@hh.se Accepted for publication: 24 March 2009 doi:10.1111/j.1365-2753.2009.01233.x

Abstract
Rationale and aim There is a lack of evidence that nursing care plans affect patient outcomes. The aim of the present study was to evaluate whether documented nursing care plans affect patient outcomes by the use of nursing-sensitive outcome indicators. Method A retrospective, cross-sectional design with patients surveyed at time of discharge from two hospital units for patients with stroke and after 23 weeks at home. The control unit (A) did not employ care plans. The intervention unit (B) used care plans on a daily basis. Outcome was measured by the National Stroke Register and the two questionnaires: Quality of patients perspective and Euroquol ve dimensions. Data were collected over a 5-month period (October 2007February 2008). Results In total, 87 patients were included in the study. Patient characteristics in the two groups at admission and the health-related quality of life 2 to 3 weeks after discharge were similar. The patients from unit B were more satised with individual care (P = 0.03) and participation (P = 0.007). Unit B also had a shorter length of stay (P = 0.004). Conclusions The effects of documented care plans are difcult to evaluate from a patient perspective, as many factors can improve the outcome. However, documentation of care plans is a prerequisite for measuring the quality and outcome of the care provided. This small study indicates that documented care plans may affect patient satisfaction regarding individual care, participation and length of hospital stay.

Introduction
According to an extensive literature review [1] focusing on studies of nursing diagnosis and nursing care, there is a lack of evidence that documented nursing care plans affect patient outcomes. However, the review revealed that the documentation of nursing diagnoses improved the quality of documented patient assessments and identication of commonly occurring diagnoses within similar settings. Escalating health care costs along with a population increasingly consisting of old people demand costeffectiveness and quality of care in the health care system. This means that professionals need to employ interventions that lead to the best outcome in order to ensure effective use of resources. Nurses are educated to make clinical judgments about an individuals response to actual and potential health problems/life processes, state the expected outcome of nursing care and formulate a plan for nursing care interventions. However, the part of the nursing process that includes nursing diagnosis, outcome and

interventions as set out in a care plan is seldom documented in clinical settings [2,3]. The documentation of nursing care can be seen as an indicator of quality, as its ultimate aim is that it should serve as a tool to achieve high-quality care as well as guaranteeing continuity and safety during hospitalization in addition to facilitating the use of evidence-based care [4]. According to a study by Bjrvell et al. [5], documentation of nursing diagnoses obliges nurses to focus on nursing care. After an intervention, the nurses found that a structured way of documenting nursing care made them change from focusing on medical technical aspects to a greater emphasis on nursing competence. Axelsson et al. [6] also found that nurses perceived that documenting nursing diagnoses helped them to view the patient as an individual as well as a whole human being, increased their awareness of nursing practice and had a positive inuence on their professional role. The study concluded that nurses were motivated to document nursing diagnoses by the wish to more deeply analyse the patients nursing needs. From the above, it can be concluded
611

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Journal of Evaluation in Clinical Practice 16 (2010) 611618

Evaluation of documented nursing care plans

I. Jansson et al.

that the documentation of care plans has an impact on the quality of care, although there is a lack of evidence as to its outcome from a patient perspective.

Background
Internationally, six prerequisites have been dened as essential for high-quality care: knowledge-based and adapted to its purpose, safe, patient focused, efcient, equal and provided within a reasonable timeframe [7]. Nursing-sensitive outcome indicators (NSOI) are described by the American Nurses Association (ANA) as focusing on how patients, and their conditions, are affected by their interaction with nursing staff [8, p. viii]. A literature review [9] stated that the outcomes generally considered as sensitive to nursing care can be classied into the following four NSOI groups: Functional, which includes physical and psychosocial functioning in addition to selfcare abilities, Clinical, which comprises symptom control or management, Perceptual, which includes satisfaction with nursing care and with care outcome and Safety, which consists of adverse incidents and complications, such as thrombosis, falls, bleeding and infection. As a result of an inventory of NSOI, the International Council of Nursing (ICN) [10] added the NSOI group use of health care, comprising the following three indicators: length of hospital stay because of increased morbidity, unplanned emergency department visits and hospital re-admission. Nursing-sensitive outcome indicators have been used in different studies to measure the effect of nursing care. Following an extensive literature review on the state of the art in the area of stafng and patient safety, Blegen [11] stated that patient safety associated with nursing care is most visible in hospital inpatient units. The most frequent NSOI in the review linked to stafng was pneumonia, followed by failure to save life, urinary tract infections, mortality as well as falls and injuries such as pressure ulcers. Other NSOI that were found to be linked to stafng were length of stay (LOS), postoperative infections and thrombosis. NSOI were also used in a quality project by ANA, the aim of which was to examine patient care outcomes in acute care hospitals in California. ANA performed a pilot study to assess the feasibility of measuring NSOI [12], after which they started to develop the California Nursing Outcomes Coalition (calNOC) database [13]. The data include hospital nurse stafng, patient days, patient falls, pressure ulcers and restraint prevalence, registered nurse education and patient satisfaction with nursing care. Additional indicators under development include standardized nursing assessment and documentation of patients activities of daily living (ADL) in addition to measurement of nursing-sensitive medication errors. Johnson et al. [14] also describe the development of a tool for reporting clinical nursing data, The nursing quality indicator report, divided into four sections: human resources, patient satisfaction, infection control and clinical nursing indicators such as falls, pressure ulcers, nursing documentation and medical errors. Quality work based on the results from the data has led to a reduction in the number of patient falls, improved nursing documentation and fewer medication errors [14]. In order to develop the quality of care and adopt a more evidence-based approach, it is necessary for nurses to employ the overview of knowledge that is available in the national guidelines and compilations of evidence in their respective countries, for
612

example The Cochrane Collaboration [15]. In Sweden, the National Board of Health and Welfare [7] states that systematic quality work should be pursued in order to satisfy the demands of patient safety, patient satisfaction and cost-effectiveness. National guidelines should form the basis for the development of institutional practice guidelines that describe responsibility, goals, guiding principles, routines and treatment procedures. As a nal step, it should be possible to follow the care process for each individual patient by means of the documentation. This means that nurses documentation of standard or individual care plans should proceed from a description of the patients care needs, problems or risks in the form of a care diagnosis, as well as stating interventions and expected outcome. These parts of the care plan can either be written in the authors own words, which is usual in Sweden or in standardized language such as that employed for diagnosis; North American Nursing Diagnosis Association (NANDA), interventions; nursing interventions classications (NIC) and outcome; nursing outcome classication (NOC) [16]. The aim of the present study was to evaluate whether documented nursing care plans affect patient outcomes by the use of NSOI.

Method
Design
The retrospective, cross-sectional design took the form of a study with patients surveyed at the time of discharge from the stroke unit and 23 weeks after going home. The study compared two hospital units treating patients with stroke at two different hospitals located in a rural part of Sweden. Although the patients at the two units were not matched, the inclusion criteria were identical, and the medical care, as well as routines and conditions in the two settings, was as similar as possible. The control unit (A) did not document care plans. The intervention unit (B) documented all steps of the nursing process including nursing care plans. In this study nursing care plans mean either a standardized care plan dened by Carpenito [17] as a printed plan that describes the nursing care to be provided for a patient, family or a group, or a individualized care plan dened as a written plan following the steps of nursing process describing the nursing care to be provided for the individual patient. In unit B both these types of care plans were used as a daily routine.

Settings
We selected two hospitals in order to nd a ward that did not document care plans (unit A) and one that did so (unit B), because of the fact that our inquiries have shown that nurses within the same hospital clinic work in a similar manner. The two hospitals were located within the same council area, which means that they were subject to the same political control and allocation of resources. The only difference was that unit A carried out all rehabilitation interventions within the ward, while unit B sent patients requiring a longer period of care to a specic rehabilitation department in the hospital. Details about staff and routines were provided by a specialist clinical nurse in each unit. The units were based on stroke teams comprising doctors, physiotherapists and occupational therapists who worked with the patients on a daily basis. The teams also included a speech therapist, social

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

I. Jansson et al.

Evaluation of documented nursing care plans

welfare ofcer and psychologist, while a dietician and dental hygienist were included when required. Unit A had developed institutional practice guidelines for stroke, a local adaptation of the national care guidelines, but no monitoring took place of whether these guidelines were applied on the basis of each patients needs, as no individual or standardized care plans were documented in the patient records. The electronic nursing documentation system in place at the unit only made it possible to document the assessment, as there was no space in which to document individual or standardized care plans. In unit B individual care plans were documented, usually including: ADL activities, movement disabilities, nutrition, social planning, pain and ulcers. Standardized care plans were prepared in relation to knowledge, risk of falls, nutrition, enteral probe, social planning and activities which could be included in the individual patients record, when necessary. In this unit the above information was documented manually, which made it possible for the nurses to document both individual care plans and standardized care plans. The nursing diagnosis, outcome and interventions were written without a specic format; thus, no classication system such as NANDA, NIC or NOC was used. Relevant data regarding the two hospital units are presented in Table 1.

were: children younger than 16 years, death during the hospital stay, a hospital stay of less than 24 hours, inability to understand the Swedish language and a health condition that made participation inappropriate (assessed by the nurse in charge). The two main reasons for exclusion were cognitive failure and death. A total of 87 patients, 55 men and 32 women, were consecutively included. A questionnaire was sent to the patients home address 2 to 3 weeks after discharge, which was completed and returned by a total of 65 participants, 41 men and 24 women (See Fig. 1). No power calculation was performed, as it was regarded as a small study aimed at investigating the possibility of evaluating whether the documentation of care plans has an impact on patient outcomes and, if so, which measurement instruments are relevant.

Data collection
A nurse in each ward was selected to be responsible for obtaining consent from the patients who agreed to participate in the study as

Table 1 Data pertaining to the two hospital units Unit A Unit B Yes Yes 25 Day 1 4 3 5 3

Participants
Patients with suspected stroke were admitted to the two participating units. When the diagnosis of stroke (trombosis or bleeding) was conrmed, the patients were entered into the National Stroke Register (NSR) [18]. During the study period, a total of 231 patients at the two units were registered in the NSR. Inclusion criteria were: patients admitted to the unit and diagnosed as suffering from stroke during the study period. Exclusion criteria

Standardized care plans Individualized care plans Patients Staff Stroke nurse, weekdays Nurses, weekdays Nurses, weekend Assistant nurses, weekdays Assistant nurses, weekend

No No 29 Day 1 4 3 5 3

Evening 3 3 3 3

Evening 3 3 3 3

Admitted with diagnosed stroke, units A and B N = 231

At time of discharge Unit A Data collection: National stroke register n = 52 Male 63% (n = 33) Female 37% (n = 19)

At time of discharge Unit B Data collection: National stroke register n = 35 Male 63% (n = 22) Female 37% (n = 13)

Exclusion criteria <16 years LOS < 24 hours Inability to understand the Swedish language Health condition too poor (mainly cognitive failure) Death

Figure 1 Data collection and inclusion of patients.

2-3 weeks after discharge Unit A Questionnaires returned n = 36 Male 61% (n = 22) Female 37% (n = 14) Completed without assistance 81% (n = 29) With help from relatives 14% (n = 5) Relatives only (n = 1)

2-3 weeks after discharge Unit B Questionnaires returned n = 29 Male 66% (n = 19) Female 34% (n = 10) Completed without assistance 62% (n = 18) With help from relatives 31% (n = 9)

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

613

Evaluation of documented nursing care plans

I. Jansson et al.

well as for registering them, which took place during the period October 2007February 2008. Each NSOI group was measured as follows: Functional NSOI was studied by means of the Euroquol ve dimensions (EQ-5D). EQ-5D was chosen for measuring healthrelated quality of life (HRQOL). It is a standardized, non-diseasespecic instrument, which provides a simple descriptive prole of a persons HRQOL. The questionnaire includes ve dimensions of health: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Each dimension comprises three levels: no problems, moderate problems and severe problems [19]. Clinical and perceptual NSOI was measured by the Quality of patients perspective (QPP) instrument. The QPP was chosen in order to assess patients self-reported satisfaction with care. Each question is evaluated by the respondent in two ways, the rst being a factual assessment, perceived reality, and the second an estimation of its subjective importance based on a 4-point Likert-type scale. The instrument was designed to be analysed by item as well as to be combined into a score [20]. Thirteen questions were selected from the basic instrument in order to measure satisfaction with nursing care; one question about pain control, one pertaining to the general atmosphere on the ward, two on the subject of education and information, three on the opportunity for participation in the care as well as whether individual care was provided, two about physical needs and four about the encounter with the nurses. Safety NSOI involved data from the NSR [18] about unexpected complications, that is, pneumonia, thrombosis and fractures that could be caused by falls. Data from this register also included the patients status before the stroke and at time of admission as well as LOS in hospital. Use of health care NSOI was measured by additional questions about any follow-up visits and re-admissions in the weeks following the patients return home. The NSR [18] was lled in by the nurse at the time of the patients discharge to the home. The postal questionnaire comprised the EQ-5D and QPP instruments in addition to questions about use of health care, all of which were adjusted to the medical condition of stroke or cerebral bleeding and considered easy to answer. The questionnaire was sent to the patients 2 to 3 weeks after discharge (Fig. 1).

use of health care, the chi-squared distribution was tested. Statistical signicance was considered when P < 0.05.

Validity and reliability

The QPP has been tested and found to have satisfactory validity and reliability, with Cronbachs alpha values of 0.690.93 [21]. The EQ-5D is a validated, standardized instrument that has been employed in many studies for the clinical and economic evaluation of health care [19]. The QPP has been used by Muntlin et al. [22], who stated that many of the identied areas are related to nursing care. To our knowledge, it has not been tested before whether the effect of documentation on NSOI can be measured by means of these instruments.

Results
Participant characteristics
Most of the participants from the two units were fully conscious when admitted to hospital. There were no differences in terms of age, sex or physical functioning prior to the stroke. The participants characteristics are presented in Table 2.

Functional NSOI
There were no differences between the two units in HRQOL 2 to 3 weeks after discharge. While on the whole the participants reported no problems in the areas of mobility and self-care, a number of patients experienced moderate problems related to usual activities such as work, household chores, family and leisure activities as well as pain/discomfort and anxiety/depression (Table 3).

Clinical and Perceptual NSOI

The patients in both units reported satisfaction with nursing care by means of a four point scale from 1 (do not agree at all) to 4 (fully agree). They also had the opportunity to tick not applicable, which means that the number of patients who answered each question varies, as these responses were not included. The patients from unit B reported a higher level of satisfaction: I had good opportunity to participate in decision that applied to my care (P = 0.007) (question 8) and My care was determined by my own requests and needs rather than the staffs procedures (P = 0.03) (question 9). No differences were observed in the patients assessment of the subjective importance of the different questions (Table 4).

Ethical considerations
Permission to carry out the study was granted by a Regional Ethics Committee. The participants were provided with verbal and written information, after which the nurse obtained their informed written consent. Participation was voluntary and the informants were told that they could withdraw from the study at any time. Condentiality was guaranteed.

Safety NSOI
Unexpected complications during the hospital stay such as the occurrence of fractures, thromboses and pneumonia were measured. The prevalence in unit A was one fracture and one case of pneumonia, while there were no reports of any complications in unit B. The low level of complications means that no differences can be measured between the units.

Data analysis
The data were mainly at a nominal and ordinal level in order to generate descriptive statistics or require non-parametric analyses. The MannWhitney U-test was employed to analyse differences in the QPP and EQ-5D between two unpaired groups (units A and B). In order to compare two sets of nominal data, that is, age, sex and
614

Use of health care NSOI

The LOS differed between the two units, with unit B having signicantly shorter care times (P = 0.004). The number of visits

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

I. Jansson et al.

Evaluation of documented nursing care plans

Table 2 Participant characteristics, from the National Stroke Register

data Participants Male Female Age, mean value Diagnosis: bleeding Diagnosis: infarction Degree of consciousness on admission RLS 1 fully awake RLS 23 partly unconscious RLS 48 unconscious Type of residence before the illness Own home Own home with home-help service Municipal special housing Mobility before the illness Fully mobile Fully mobile indoors Required assistance with mobility Toilet visits before the illness Managed visits to the toilet without assistance Required assistance with visits to the toilet Dressing before the illness Managed without assistance Required assistance

Unit A 52 63% (n = 33) 37% (n = 19) 74 years 4% (n = 2) 96% (n = 50) 96% (n = 50) 2% (n = 1) 0% (n = 0) 92% (n = 48) 2% (n = 1) 6% (n = 3) 96% (n = 50) 2% (n = 1) (n = 0) 98% (n = 51) 2% (n = 1) 98% (n = 51) 2% (n = 1)

Unit B 35 63% (n = 22) 37% (n = 13) 74 years 14% (n = 5) 86% (n = 30) 100% (n = 35)

89% (n = 31) 11% (n = 4) (n = 0) 100% (n = 35) (n = 0) (n = 0) 100% (n = 35) (n = 0) 100% (n = 35) (n = 0)

No statistical differences were found between the two units.

Table 3 Euroquol ve dimensions, health-related quality of life Unit A Moderate problems 26% 20% 37% 67% 41% (n = 9) (n = 7) (n = 12) (n = 22) (n = 13) Severe problems 3% (n = 1) 3% (n = 1) 12% (n = 4) 6% (n = 2) 6% (n = 2) Unit B Moderate problems 36% 18% 50% 52% 39% (n = 10) (n = 5) (n = 14) (n = 14) (n = 11) Severe problems 0 0 0 0 0

Dimensions Mobility Self-care Usual activities Pain/discomfort Anxiety/depression

n 34 34 32 33 32

No problems 71% 76% 50% 27% 50% (n = 24) (n = 26) (n = 16) (n = 9) (n = 17)

n 28 28 28 27 28

No problems 64% 82% 50% 48% 61% (n = 18) (n = 23) (n = 14) (n = 13) (n = 179)

P 0.511 0.623 0.133 0.139 0.385

Table 4 Quality of care from the patient perspective Quality of care, perceived reality Unit A Statement 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. I received useful information on how examinations and treatments would take place I received useful information on self-care; how I should take care of myself The nurses and assistant nurses seemed to understand how I experienced my situation The nurses and assistant nurses were respectful towards me The nurses and assistant nurses showed commitment; cared about me I received effective pain relief I talked to the nurses in private when I wanted to I had good opportunity to participate in decision that applied to my care My care was determined by my own requests and needs rather than the staffs procedures My relatives and friends were treated well There was a pleasant atmosphere on the ward I received food and drink that I like I received the best possible physical care, e.g. help with my personal hygiene n 34 28 35 32 30 18 21 26 31 32 34 34 20 M 3.38 2.82 3.66 3.72 3.83 3.28 3 2.65 3.1 3.63 3.44 3.38 3.1 SD 0.74 1.09 0.68 0.58 0.38 0.89 0.95 0.94 0.7 0.71 0.79 0.78 0.91 Unit B n 29 26 27 27 28 9 17 24 27 25 29 29 18 M 3.41 3.35 3.67 3.7 3.71 3.44 3.24 3.38 3.48 3.56 3.59 3.55 3.39 SD 0.73 0.94 0.62 0.61 0.6 1.01 0.97 0.88 0.75 0.82 0.63 0.87 0.85 P 0.87 0.06 0.87 0.96 0.57 0.47 0.34 0.007 0.03 0.8 0.52 0.21 0.27

Bold values are statistically signicant.

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

615

Evaluation of documented nursing care plans

I. Jansson et al.

n LOS, average Planned follow-up visits Hospital Outpatient clinic Emergency visits Hospital Outpatient clinic Discharged to Own home Own home with rehabilitation Municipal accommodation Municipal accommodation with rehabilitation/geriatric rehabilitation LOS, length of stay; n.s., not signicant. 52 34 34 33 33 52 52 52 52 52

Unit A 12.46 days 12% (n = 4) 44% (n = 15) 15% (n = 5) 18% (n = 6) 75% (n = 39) 6% (n = 3) 10% (n = 5) 6% (n = 3) 4% (n = 2)

n 35 26 26 26 26 33 33 33 33 33

Unit B 6.26 days 27% (n = 7) 31% (n = 8) 8% (n = 2) 4% (n = 1) 70% (n = 23) 10% (n = 3) 0 0 (21% n = 7)

P 0.004 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s.

Table 5 Use of health care

to the outpatient clinic or hospital within the rst 23 weeks after discharge did not differ in terms of scheduled or emergency visits (Table 5).

Discussion
Study limitations
In this study we wished to investigate whether or not nursing care based on a structured method of documenting care plans affects patient outcomes. Two questions emerged at the start of the planning of the study: which variables are useful for measuring patient-related effects? Is it possible to measure whether the documentation of care plans affects patient outcomes? Because of the fact that we did not know which variables might be affected by the documentation of care plans, we decided to measure as many NSOI as possible based on existing documentation in the Stroke Register, in addition to the questions asked in the questionnaires. We chose to send the questionnaires to the patients 2 to 3 weeks after their discharge to the home, as it was assumed that by then they had had time to digest their experiences and were no longer in a state of dependence vis vis the health care professionals. A longer period was assumed to constitute a greater risk of the emergence of new nursing problems that were not an issue when the care planning was performed at the hospital. This interval can nevertheless have led to patients forgetting a great deal of how they experienced the care. We only obtained information about HRQOL at that particular point in time. It might have been of greater interest to explore HRQOL at the moment of discharge to the home since the measure of HRQOL then presumably had been more related to the care and hospital stay. Further, HRQOL at home might be affected by the fact that it was good to be back home. We could also have employed a more sensitive instrument for measuring ADL, as in the study by Glasson et al. [23] where the Bartels ADL-index was used to evaluate a new care model. Nevertheless, a remaining problem is the difculty in determining which interventions had an effect on the results. We are aware of the difculties of drawing conclusions from a study comparing two different sites using and not using documented care plans, because, although efforts have been made to
616

ensure that they are otherwise as similar as possible, any number of factors could contribute to any observed differences. However, we argue that studies of these kinds are needed to further explore the value of care plans and the impact of NSOI, in spite of the many difculties. One notable difference between the units was that unit A had patients staying for long-term rehabilitation. When excluding the few patients who had stayed for long-term rehabilitation, the difference in LOS was still signicant. According to a review by Blegen [11], the aim of previous studies was often to explore whether the number of registered nurses at the unit has an impact on the NSOI. In our study, although strive for equal conditions in the two ward, unit A had four more beds that could mean in general a heavier workload at this unit. However, we have made the assumption that four more patients might not affect quality of care to the amount that it has an impact on NSOI and instead have drawn a very cautious conclusion that other factors such as working methods also may have inuenced the outcome. While we know that in unit B the nurses working method involved documenting all the steps in the nursing process, we do not know whether the nurses in unit A worked along those lines, although they did not actually document the procedure. We also have no knowledge pertaining to the quality of the documentation produced by unit B. A previous study [24] revealed that nursing diagnoses are often non-specic and inaccurately formulated, leading to inappropriate nursing goals and inadequately dened nursing interventions. The nurses in unit B who documented care plans did not employ the classication system for diagnosis, outcome and interventions. A previous study [25] demonstrated that the implementation of the classication systems NANDA, NIC and NOC led to a higher quality of nursing diagnosis documentation as well as etiology-specic nursing interventions and outcomes. Nevertheless, studies [5,6] have revealed that even when the steps are documented without a specic format, nurses exhibit a higher level of nursing expertise orientation; thus, for this reason, we wished to investigate whether the documentation of care plans has an impact on NSOI. The NSOI recognized by the literature can also be inuenced by many other factors, for example, the organization of care in terms of structure and available resources, the work of other professional

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

I. Jansson et al.

Evaluation of documented nursing care plans

categories, the well-being of patients and their family members, previous experiences of care, etc. In order to minimize the risk of organizational changes in the units, we wished the data collection period to be as short as possible, which contributed to the small number of participants. Nevertheless, certain measurable differences emerged between the units, which may have been more obvious had the study been conducted with a larger number of participants. Another limitation may be the fact that patients who perhaps have the greatest need of care planning, that is, those with a very poor health status, were excluded from this study because of cognitive failure, which meant that they were unable to give their informed consent and complete the questionnaire. More patients in unit B had assistance from a next of kin when lling in the questionnaires than in unit A, which might reect the fact that patients in unit B were in a worse condition. However, when comparing data from the NSR, this assumption was not conrmed. Finally, it is difcult to tell whether or not the answers from unit B reect opinions from the next of kin to a greater extent than in unit A.

individualized care compared with cases where they are not employed. Earlier studies [27,28] demonstrated that patients do not perceive themselves as involved in the care that they receive in the ward and prefer to play a passive role. In this present study we do not know how the answer is in this specic question. The fact that more patients in unit B reported that: I was given a good opportunity to participate in decisions that applied to my care could be an indication that care plans allow patients to become more involved in their care. In further studies it might be interesting also to review the nursing documentation regarding both quality and quantity of recorded nursing interventions. According to the ICNs code of ethics for nurses [29], respect for human rights, including cultural rights, the right to life and choice, to dignity and to be treated with respect is inherent in nursing. The nurses in both units appear to have succeeded in communicating this attitude to the patients, as the latter were very satised at being treated with respect and at the nurses paying them attention. This aspect of nursing thus functioned satisfactorily, irrespective of whether or not care plans were documented.

Discussion of ndings
Safety NSOI Functional NSOI
Usual activities pain/discomfort and anxiety/depression during the 23 weeks following discharge constituted problems for the patients in both groups. According to a previous study [1], pain is the most common nursing diagnosis, and participating nurses from unit B stated that they often produced individual care plans on this basis. Despite this, no difference was found between the units in this respect. Nor did the question in the QPP pertaining to pain reveal any differences in how patients perceived the pain relief medication received during their hospital stay. With regard to activities, the participating nurses from unit B stated that they documented both individual and standardized care plans. However, despite this, we were unable to observe any differences in this area, which may be due to the care plans having no effect on nursing care, that the study population was too small or that the instrument employed was not sensitive enough. The patients who participated in the study suffered few complications during their time in hospital in the form of fractures, thromboses and pneumonia. It would have been interesting to measure whether the incidence of these complications was greater among the most seriously ill patients with the largest care need who were excluded from the present study. As the nurses in unit B documented care plans pertaining to the risk of falls as well as to activities where the possible outcome could be fractures or thromboses, it would be interesting to measure these indicators in a larger study where differences might occur because of the use of care plans. It would also have been interesting to measure the incidence of falls and pneumonia at the units, the latter being the most common NSOI according to Blegen [11].

Use of health care NSOI

The large difference in length of hospital stay is noteworthy. Although unit B, in which care plans were documented, had signicantly shorter care times, it is not possible to infer that this is due to structured nursing care based on care plans. An explanation may be the fact that the patients in unit A also remained in the ward for long-term rehabilitation, while the patients in unit B were transferred to a rehabilitation department if their need for rehabilitation was estimated to be extensive. A major problem described in studies measuring NSOI [11,13] is to ensure that nurses employ the correct documentation in relation to the various indicators. Blegen [11] states that we need standardized data of patient outcomes in order to conduct more rigorous studies than in the past. Muller-Staub [2] concludes that standardized terms in the documentation make it easier to measure the quality of nursing. Nursing documentation can be seen as one form of NSOI [4] and this becomes particularly obvious when we try to measure the effect of something that is not clearly documented. Care plans could result in improved quality of care as well as constitute a measure of high-quality care.
617

Clinical and perceptual NSOI

A study by Wilde Larsson et al. [26] revealed that the patients perception of the quality of care is inuenced by his or her level of HRQOL. The fact that the HRQOL as measured by the EQ-5D was identical at the two units strengthens the result of the instrument that measures quality from a patient perspective (QPP). In Axelssons study [6], the nurses perceived that nursing diagnosis claried the patients individual needs. In our results more patients in unit B stated: My care was carried out in accordance with my own requests and needs rather than staff procedures, which may indicate that the patients cared for in unit B where nursing diagnoses were employed, also perceived that the care was guided by their individual needs to a higher degree than in unit A, where nursing diagnoses were not used. A previous study [24] revealed that nursing diagnoses do not always agree with the patients perception of his or her problem and needs, although the present study indicates that such diagnoses lead to more

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

Evaluation of documented nursing care plans

I. Jansson et al.

Conclusions
It is difcult to evaluate whether documented care plans affect patient outcomes, as there are many other factors that have an impact. Nevertheless, the present small study revealed that patients who were cared for on the basis of documented care plans perceived not only that the care was more individual, that they had a greater opportunity to participate in decision making, but also that the use of care plans had no inuence on patients experienced HRQOL, prevalence of unexpected complications or use of health care after hospital discharge. We were unable to establish whether care plans affect patient outcome but conclude that nurses need to document care plans comprising nursing diagnosis, interventions and outcome in order to be able to measure the quality of the nursing care provided. The use of standardized terms in this documentation would further facilitate the measurement of quality of care. A larger study investigating how documentation affects patient outcome could, in conjunction with an intervention comprising the implementation of documentation of diagnosis, interventions and outcome, also measure NSOI both before and after the implementation.

Acknowledgements
This study was performed by valuable assistance from one clinical nurse specialist at each stroke unit as well as from Kalle Svensson and Tina Rinaldo at Improve IT, Halmstad. We also gratefully acknowledge Assistant Professor Jos Ferraz Nunes for advice regarding the design and Assistant Professor Monika Fagevik Olsn for advice regarding statistical analysis.

References
1. Muller-Staub, M., Lavin, M. A., Needman, I. & Achterberg, T. (2006) Nursing diagnosis, interventions and outcomes application and impact on nursing practice: systematic review. Journal of Advanced Nursing, 56 (5), 514531. 2. Ehrenberg, A. & Birgersson, C. (2003) Nursing documentation of leg ulcers: adherence to clinical guidelines in a Swedish primary health care district. Scandinavian Journal of Caring Sciences, 17 (3), 278 284. 3. Trnvall, E., Wahren, L. K. & Wilhelmsson, S. (2007) Impact of primary care management on nursing documentation. Journal of Nursing Management, 15 (6), 634642. 4. Ehnfors, M. & Ehrenberg, A. (2007) Quality indicators for nursing documentation in the patient record. In Quality Indicators in Nursing (ed. E. Idvall), pp. 3039. Stockholm: Gothia frlag. [in Swedish]. 5. Bjrvell, C., Wredling, R. & Thorell-Ekstrand, I. (2003) Improving documentation using a nursing model. Journal of Advanced Nursing, 43 (4), 402410. 6. Axelsson, L., Bjrvell, C., Mattiasson, A-C. & Randers, I. (2006) Swedish registered nurses incentive to use nursing diagnosis in clinical practice. Journal of Clinical Nursing, 15 (8), 936945. 7. The National Board of Health and Welfare (2006) Good care about management systems for quality and patient safety in health care. [in Swedish]. ISBN: 91-85482-05-6. 8. American Nurses Association (1995) Nursing Report Card for Acute Care. Washington DC: American Nurses Publishing.

9. Doran, D. (ed.) (2003) Nursing Sensitive Outcomes: State of the Science. London: Jones and Bartlett Publishers Inc. 10. I. C. N. International Council of Nurses (2008) Nursing matters fact sheet: nursing sensitive outcome. Available at: http://www.icn.ch/ matters_indicators.htm (last accessed 28 May 2008). 11. Blegen, M. A. (2006) Patient safety in hospital acute care units. Annual Review of Nursing Research, 24, 103125. 12. Langemo, D. K., Anderson, J. & Volden, C. M. (2002) Nursing quality outcome indicators The North Dakota study. Journal of Nursing Administration, 32 (2), 98105. 13. Aydin, E., Burnes Bolton, L., Donaldsson, N., Storer Brown, D., Buffum, M., Elashoff, J. D. & Sandhu, M. (2004) Creating and analyzing a statewide nursing quality measurement database. Journal of Nursing Scholarship, 36 (4), 371378. 14. Johnson, K., Hallsey, D., Meredith, R. & Warden, E. (2006) A nurse driven system for improving patient quality outcomes. Journal of Nursing Care Quality, 21 (2), 168175. 15. The Cochrane Collaboration (2008) Available at: http://www. cochrane.org/ (last accessed 8 September 2008). 16. Johnson, M. (2006) NANDA, NOC and NIC Linkages. Nursing Diagnosis, Outcomes and Interventions. St Louis, MO: Mosby. 17. Carpenito, L. J. (2000) Nursing Diagnosis Application to Clinical Practice, 8th Edn. Philadelphia, PA: Lippincott. 18. Riks-stroke (2008) Available at: http://www.riks-stroke.org/index.php (last accessed 3 June 2008). 19. Euroqual (2008) Available at: http://www.euroqol.org/ (last accessed 3 June 2008). 20. Wilde Larsson, B., Larsson, G., Larsson, M. & Starrin, B. (2001) The QPP Book, a Guide to the Questionnaire QPP, Quality from a Patient Perspective, 4th Edn. Stockholm: Vrdfrbundet. [in Swedish]. 21. Wilde Larsson, B. (2000) Does the method of data collection affect patients evaluations of quality of care? International Journal of Nursing Practice, 6, 284291. 22. Muntlin , Gunningberg, L. & Carlsson, M. (2005) Patients perception of quality of care at an emergency department and identication of areas for quality improvement. Journal of Clinical Nursing, 15 (8), 10451056. 23. Glasson, J., Chang, E., Chenoweth, L., Hancock, K., Hall, T., HillMurray, F. & Collier, L. (2006) Evaluation of a model of nursing care for older patients using participatory action research in an acute medical ward. Journal of Clinical Nursing, 15 (5), 588598. 24. Lunney, M. (2003) Critical thinking and accuracy of nursing diagnosis. International Journal of Nursing Terminologies and Classications, 14 (3), 96107. 25. Muller-Staub, M., Needman, I., Odenbreit, M., Lavin, M. A. & Achterberg, T. (2007) Improved quality of nursing documentation: Results of a Nursing diagnosis, Interventions, and Outcomes implementation study. International Journal of Nursing Terminologies and Classications, 18 (1), 517. 26. Wilde, B., Starrin, B., Larsson, G. & Larsson, M. (1993) Quality of care from a patient perspective: a grounded theory study. Scandinavian Journal of Caring Sciences, 7 (2), 113120. 27. Doherty, C. (2005) Patients preferences for involvement in clinical decision-making within secondary care and the factors that inuence their preferences. Journal of Nursing Management, 13 (2), 119127. 28. Florin, J., Ehrenberg, A. & Ehnfors, M. (2005) Patients and nurses perceptions of nursing problems in an acute care setting. Journal of Advanced Nursing, 51 (2), 140149. 29. I. C. N. International Council of Nurses (2005) The ICN Code of Ethics for nurses. Available at: http://www.icn.ch/icncode.pdf (last accessed 25 February 2008).

618

2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd

Copyright of Journal of Evaluation in Clinical Practice is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.