Pressure ulcers continue to be a common adverse event among patients in hospitals and in community settings, causing both suffering1 and increased costs in the healthcare system.2 The Institute for Healthcare Improvement in the United States has added pressure ulcers as one of 12 quality indicators to reach the campaign goal of saving 5 million lives from medical harm. The specific goal for pressure ulcers is to reduce the incidence of hospital-acquired ulcers to zero by the end of 2008.3 Today’s healthcare systems are complex, with many care givers and professional groups. A pressure ulcer can develop—for example, at home, or in the ambulance, the emergency department, the radiology department, the operating room, the ward or during rehabilitation. Old and immobile patients, as well as those with severe acute illness or neurological deficit, are at risk for developing pressure ulcers.45 Prevention and treatment of pressure ulcers require a multidisciplinary approach, involving physicians, nurses, physiotherapists, occupational therapists and dieticians.67

Varying prevalence of pressure ulcers has been reported from different countries, ranging from 10.1% to 23.1%.48^–10 Because of the lack of standardised methods for determining incidence and prevalence rates, these studies are not always comparable. To address this problem, the European Pressure Ulcer Advisory Panel (EPUAP) has developed a methodology to ensure the validity of prevalence surveys throughout Europe.1112 This methodology has been tested in 25 hospitals in five European countries, with the conclusion that the methodology is sufficiently robust to measure and compare pressure-ulcer prevalence in different countries. The pressure-ulcer prevalence in Europe was higher than expected (18.1% including grade 1–4). Relatively few patients received pressure-reducing mattresses, which is recommended by a Cochrane review5 or were repositioned regularly.

The EPUAP prevalence study was conducted in Sweden for the first time in 2002.13 In total, 612 patients were surveyed in a university hospital, revealing an overall pressure-ulcer prevalence of 23.9%. Departments with the highest prevalence rates were geriatric (59.3%), medicine (23.6%) and surgery/orthopaedic (26.8%).

The electronic health record (EHR) has the potential to support the use of research-based clinical guidelines by integrating templates for decision support.14 Reviews have shown that the use of computer-based decision support enhances the performance of physicians.15 Although there are few studies on the impact of such support on nursing performance, there are indications of improved nurse performance and clinical judgement.16 In the university hospital, where this study took place, the EHR was broadly implemented for all professionals in 2004–2006. According to Swedish laws and regulations, a patient problem should be identified, and adequate intervention should be planned, implemented and evaluated, as well as documented.1718 Assessments should not be documented daily if nothing new has occurred. Pressure ulcers were also identified as a quality indicator at hospital level to be reported annually by all departments, and templates to support recording of pressure-ulcer data were integrated in the EHR. As prevalence surveys are time- and labour-consuming, such templates can provide a basis for standardised recording and regular feedback quickly and easily, on a monthly basis, to the staff on the wards. This would empower the professionals to set goals and test changes for improvement.19 The EHR could be utilised in this process, provided that the data contained in the documentation are reliable.

The aim of the study was to compare the accuracy in recording of pressure-ulcer prevalence and prevention in hospital care before and after implementing an EHR with templates for pressure-ulcer assessment.

METHODS

Sample

All inpatients at the departments of geriatrics, surgery, orthopaedics and medicine of a Swedish university hospital were inspected for the presence of pressure ulcers during 1 day in 2002 (n = 357), and the same prevalence study was repeated 4 years later, during 1 day in 2006 (n = 343). In the present study, surgery/orthopaedics will be defined as the “surgical department.” The records of all patients were audited retrospectively in 2002 (n = 413) and in 2006 (n = 343) for the presence of documentation on pressure ulcers. As there was no identification of individual patients in 2002, the record audit was based on the total number of records for inpatients registered for the day of the pressure-ulcer survey. Thus, more records were included compared with the number of patients, as some patients were not available in their beds or on the ward at the time of the survey. Data from the records were retrieved from a retrospective period of maximum 3 months before the physical examination of the patients. There were 19 (4.4%) missing records in 2002 and five (1.4%) in 2006. Table 1 shows the patient characteristics for the samples. In 2002, it was not possible to match data on an individual basis from the physical examination with audit data.

Instruments

The EPUAP prevalence methodology was used for the physical examination of the patients.1112 Information on risk assessment, pressure ulcers and prevention was gathered on a one-page data-collection form. The form also included patient-demographic data, such as age, gender, and department. Pressure ulcers were assessed as Grade I, non-blanchable erythema of intact skin; Grade II, partial-thickness skin loss involving epidermis, dermis or both; Grade III, full-thickness skin loss involving damage to or necrosis of subcutaneous tissue that may extend down to, but not through, underlying fascia; and Grade IV, full-thickness skin loss with extensive destruction, tissue necrosis or damage to muscle, bone or supporting structures.20 The following preventive strategies were recorded; the use of pressure-reducing mattresses, chair-cushions and planned repositioning in bed and chair.

An audit instrument for review of patient records contents regarding pressure ulcers was developed by the researchers for the study based on the EPUAP protocol.21

Procedure

All patients received verbal and written information about the study. Each patient was visited by two nurses, that is a specially trained data collector (non-ward nurse) and a staff nurse (ward nurse) who assessed all patients for pressure ulcers during 1 day in 2002 and 2006. The data collectors and staff nurses attended a half-day seminar to outline the aims and survey procedure, and there was specific training in pressure-ulcer classification.13

Patient records for all patients included in the pressure-ulcer survey were retrospectively identified in the hospital’s record archives in 2003 and in the EHR system in 2006. All records (physicians and nurses’ notes) were read through to identify any documentation pertaining pressure ulcers. The records holding such data were photocopied and audited by pairs of undergraduate nursing students in their final term. Before data collection, the students received training in record audit using the specific instrument (including pressure-ulcer classification).

Inter-rater reliability of assessments of the patients’ skin and record audits was calculated. Prior to the physical examination in the prevalence survey, 10 colour photos of pressure ulcers were graded by each nurse. The mean Cohen kappa was 0.82 (n = 22) in 2002 and 0.78 (n = 52) in 2006, which was judged to be an excellent agreement.22 After the inter-rater test, each photograph was discussed, and criteria for pressure-ulcer grading were presented. The research team reviewed 8.5% to 10% of the patients’ records for notes regarding pressure ulcers and preventive strategies. The Cohen kappa values varied between 0.57–1.0 (83% to 100%), which was judged to be fair to excellent agreement.22

Permission for the study was obtained from the medical directors at each department, and approval was obtained from the Research Ethics Committee of the Faculty of Medicine at Uppsala University (No. 03-012). To ensure anonymity of the patients, all personal identification in the records was removed before photocopying, and all data were treated confidentially.

Changes

After the EPUAP prevalence survey in 2002, hospitalwide interventions on pressure-ulcer care were performed, such as information to all head nurses, education for nurses, and development of guidelines for purchase and allocation of pressure-reducing mattresses. In 2004, templates for risk assessment,23 pressure-ulcer grading20 and standard care plans were developed to facilitate adequate documentation in the EHR. The EPUAP prevalence survey was repeated in 2004 with fast feedback of results. Table 2 lists the actions at the department level.

Data analysis

Accuracy of record data was judged based on the physical examination as a “golden standard.”12 As there was no identification of individual patients in 2002, the comparison was made at the group level. For 2006, we also compared data from each patient’s record with the findings from the physical examination. SPSS was used for a descriptive analysis of the data.

RESULTS

In 2002, the overall prevalence of pressure ulcers obtained by audit of paper-based patient records (n = 413) was 14.3% compared with 33.3% when two nurses examined the patients’ skin (n = 357) (table 3). The largest difference was seen in the geriatric department, where the audit revealed 22.9% pressure ulcers, and the skin inspection resulted in 59.3% of the patients assessed to have ulcers. Four years later, after implementation of the EHR, there were 20.7% recorded pressure ulcers and 30.0% found by physical examination of all inpatients (n = 343) during a 1-day survey. The accuracy of the prevalence data had changed most in the geriatric department, where the prevalence of pressure ulcers obtained by audit of the EHR was 48.1% and by physical examination 43.2%. Corresponding figures in the surgical department were 22.2% and 14.1% and in the medical department 29.9% and 10.2%. The differences between recorded data and physical examination were mainly seen in pressure ulcer grades 1 and 2. Physicians’ documentation regarding pressure ulcers was present in 18 (4.3%) of the paper-based patient records and in 28 (8.2%) of the EHR’s.

Table 4 displays that the findings of pressure ulcers from physical examination were consistent with data in the EHR in 67.9% in the geriatric department, 78.5% in the surgical department and 75.6% in the medical department, when comparing data for each individual patient.

The use of preventive strategies (pressure reducing mattresses and cushions) had increased from 27.7% to 51.6% over the 4-year period. This was not evident in the EHR, which revealed the use of pressure-reducing equipment for only 7.9% of the patients in 2006 (table 5).

DISCUSSION

After implementation of the EHR, the accuracy in recording of pressure-ulcer prevalence improved considerably when compared with a physical examination of the patients. The prevalence of pressure ulcers obtained by physical examination was 33.3% in 2002 and 30% in 2006. Less than half of the patients with pressure ulcers had any data on pressure ulcers in their paper-based records in 2002, whereas two-thirds of the patients with observed pressure ulcers in 2006 had such notes in the EHR. The combined implementation of EHR and templates for pressure-ulcer grading seems to have contributed to improved accuracy in nurses’ recording of data on pressure ulcers. However, as it was not possible to conduct the study with an experimental design, no conclusions can be made about the impact of the EHR alone.

At the departmental level, the gap between the prevalence obtained by physical examination versus audit was reduced in the geriatric department but increased in the medical department during the study period. The EHR was implemented parallel with a major re-organisation in the hospital. Kitson and co-workers have presented a conceptual framework suggesting that successful implementation of change can be explained by the relationship between evidence, context and facilitation.2425 The organisational context, including leadership and culture, is highlighted as it influences priorities and investments.26 A dedicated facilitator is important for providing motivation to initiate and support the change process. In the geriatric department, the leadership took an active interest in both pressure-ulcer care and the implementation of the EHR. They also provided a facilitator for the quality-improvement work at the departmental level, whereas the same support was not given in the surgical and medical departments.

Implementing a multidisciplinary EHR across healthcare institutions is a paradigmatic shift. The contribution to the documentation from each professional group is highlighted and questioned; there are technical problems and new ways for information exchange. There are also high expectations for the EHR to be time-saving, increase patient safety and provide useful data on quality indicators. Studies investigating the accuracy of the documentation of quality indicators in EHR are sparse. Tang et al27 found that clinically based measures derived from the EHR were more reliable compared with administrative data from billing systems, and Browne and Covington28 showed improved accuracy in risk assessments with embedded clinical indicators in the HER.

With the introduction of EHR, reports between clinicians and units will be based mainly on the documentation. If a problem is not highlighted in the patient record, there is a great risk for omission and errors in care. This implies that the under-reporting of pressure ulcers in the records may have an impact on patient safety. Patients at risk for developing pressure ulcers, or with Grade 1 ulcers, may not receive appropriate preventive measures, as their condition is not apparent in the record. Studies indicate that Grade 1 ulcers might be a more useful indicator for development of severe pressure ulcers compared with traditional risk-assessment scales.2930 The assessment in the documentation should be used by the multidisciplinary team to set goals and plan for preventive measures. The results also showed that the use of preventive equipment almost doubled over the 4-year period, even if it was not recorded. This may indicate that there was an increased awareness of pressure-ulcer prevention. The templates used in the present study have been revised, and a system for regular feedback to the wards is implemented for identified patient groups especially at risk for pressure ulcer. Thus, unnecessary documentation should be avoided and time made free for bedside care.

Methodological limitations

One limitation of the method used is that the pressure-ulcer prevalence study was a 1-day cross-sectional survey, while the record audit used a retrospective design. Grade 1 ulcers, in particular, can develop or disappear during a short period of time, depending on the patient’s condition, which explains the deviations displayed in table 3.

Conclusions

The accuracy in pressure-ulcer documentation improved in the EHR compared with the paper-based records. However, there were still deficiencies in recorded data after implementation of the EHR, which means that patient records did not serve as a valid source of information on pressure-ulcer prevalence and prevention. These findings indicate that considerable improvement can be achieved by supporting the recording of pressure ulcers with EHR and templates to facilitate and speed up the recording. Thus, we conclude that implementation of EHR and templates for standardised documentation is an important quality-improvement intervention. Considering the remaining gap between observed patient condition and recorded data, clinicians need to be cautious in the follow-up of quality indicators for pressure-ulcer care. Investigations of pressure-ulcer prevalence still require time-consuming observational surveys in order to present valid data. Researchers need to consider the issue of accuracy when using data on pressure ulcers or other quality indicators from health records.