Letters to the Editor: Prevalence and Incidence Study Sparks Issues
With much interest, we read the article, "Prevalence and Incidence Studies of Pressure Ulcers in Two Long-Term Care Facilities in Canada" (Ostomy/Wound Management, 2001;47:28-34). Studies in which both the prevalence and incidence have been examined are scarce, although they offer a great deal of information.
In their study, Davis and colleagues addressed the prevalence of pressure ulcers by all residents of the two studies and after 41 and 42 days they again measured the prevalence of pressure ulcers, but only in those residents who did not have a pressure ulcer at the first measurement. The authors characterized the latter measurement as an incidence study.
We have two comments regarding this study:
1. Prevalence is a cross-sectional count of the number of cases of a medical condition at a specific point in time. Incidence is the number of new cases occurring over a given time period. The authors operationalized the incidence as a cross-sectional count of the pressure ulcers in residents who did not have a pressure ulcer after 41 or 42 days of the first cross-sectional count. We question if this operationalization is correct for two reasons:
a. In an earlier study,1 we found that most Stage I pressure ulcers are reversible, sometimes within 1 day. This means that when measuring grade 1 - which is questionable, because it is very difficult to measure reliably - one must measure each day to detect all grade 1 ulcers. Otherwise, a risk exists of missing much of Stage I and, to a lesser degree, Stage II. In the article, incidence rates must be underestimated. What they show is that in a residential home, the prevalence of pressure ulcers is very high. During the first measurement, the prevalence rate was already high, but within 41 days, approximately 12% of those who had no pressure ulcers during the first measurement developed one or more lesions.
b. More problematic is that the authors only used residents who had no pressure ulcers during the first measurement; thus, introducing a selection bias regarding the risk of developing pressure ulcers. This dramatically influences the results. One may expect that the residents who had pressure ulcers have, in general, a higher risk of developing pressure ulcers than those who did not have a pressure ulcer at the first measurement. The authors selected residents who generally have a lower risk on developing pressure ulcers than the general population of residents, tainting the incidence rate by selection bias.
2. We question the interpretation of prevalence and incidence rates and, therefore, the value of each measurement. In comparing the two facilities, the authors conclude that the level of nursing care is equal and acceptable based on the comparable incidence rates. Even when the incidence was measured in the right way (see point 1), we do not agree with this statement. With comparable incidence rates and different prevalence rates, the difference between the two facilities may only be explained by the duration of the condition because an exact relation exists among these three quantities (prevalence = incidence x duration).2
Patients in the facility with the higher prevalence suffer longer with a pressure ulcer. This difference in duration may be attributed to any etiologic factor (medical condition) or to quality of care. Explaining the disparity of the prevalence of pressure ulcers only by a difference in case-mix is too simple and neglects the treatment effect. Incidence figures give (with a comparable population) a good insight into the quality of the preventive care. However, incidence figures do not infer anything about the quality of the treatment. Prevalence rates are a combined rate of the quality of the preventive and the treatment care. One should not draw conclusions about the individual effect of the preventive or treatment care. This is only possible when the duration of the pressure ulcers is known.
An important question still remains. Do we need prevalence measurements or do we need incidence measurements? As with most questions, one first has to identify what exactly one wants to know. To determine the effect of risk factors or to evaluate the effectiveness of specific preventive interventions, only incidence measurements are indicated. To gain insight into the extent of the problem, prevalence measurements are indicated because they provide a more complete and combined rate of both the incidence and the duration of the pressure ulcer. Prevalence rates are also easier to measure; one needs only to measure 1 day. Incidence rates must be measured daily during a specific period, which is very time-consuming.
Since 1998, four national prevalence surveys have been carried out in the Netherlands. During these surveys, the duration of the pressure ulcers also had been measured.3 This allowed a derivation of incidence figures according to the above-mentioned relations among the three quantities. Additional research4 showed that the prevalence surveys were very important in drawing attention within the institutes to the prevention and treatment of pressure ulcers. After the first survey, almost all participating institutions started to plan or implement activities to improve the prevention and treatment of pressure ulcers (developing or updating protocols, educating nurses, appointing a nurse specialist, and so on). Yearly repetition of the prevalence survey is important to continue focus on pressure ulcers and to investigate if the new activities produce a positive result.
We strongly recommend collecting data in the most expedient manner, but experience has shownthat substantial savings of time and resources may be realized in collecting prevalence data and duration and this must be emphasized.
- Ruud JG Halfens and Gerrie JJW Bours, Maastricht University, Section of Nursing Science PO BOX 616 6200 MD Maastricht The Netherlands
1. Halfens RJG, Bours GJJW, Bronner CM. The impact of assessing the prevalence of pressure ulcers on the willingness of health care institutions to plan and implement activities to reduce the prevalence. J Adv Nurs. 2001;36 (5): 617-625.
2. Freeman J, Hutchison GB. Prevalence, incidence and duration. Am J Epidemiol. 1980;112(5):707-723.
3. Bours GJJW, Halfens RJG, Huijer AH, Grol RTPM. Prevalence, prevention and treatment of pressure ulcers: descriptive study in 89 institutions in the Netherlands.” Res Nurs Health. In press.
4. Halfens RJG, Bours GJJW, van Ast JF. Relevance of the diagnosis ‘Stage 1 Pressure Ulcer’: an empirical study of the clinical course of stage 1 ulcers in acute care and long-term care hospital populations. Journal of Clinical Nursing. 2001;10:748-757 .
We accept some of the criticisms of Halfens and Bours and acknowledge that our research study was not perfect. In particular, we confirm that residents in the incidence study were not examined on a daily basis by the research teams, as was clearly stated in our paper. However, one of us was in frequent contact with nursing staff at both facilities to ensure that no cases of pressure ulcer were lost from the study due to successful treatment during the period of the incidence study. We do not believe that the incidence rate in either long-term care facility (LTCF) was underestimated due to failure to assess each resident every day of the study. As Halfens and Bours note in their letter, the measurement of incidence on a daily basis is very time-consuming and demanding on resources.
We plead not guilty to the accusation of introducing a selection bias into our incidence study of pressure ulcers. In the incidence study, we searched for the number of new cases (residents) who developed the medical condition (pressure ulcer) over a given time period (41 to 42 days) in keeping with the definition of incidence. We deliberately excluded residents who already had a pressure ulcer (which was detected at the prevalence study) because they would no longer qualify as new cases of the medical condition when they developed other pressure ulcers. Our method follows the guidelines of Frantz,1 who states that only patients who are initially free of pressure ulcer are to be included in the at-risk population when the incidence rate for pressure ulcer is determined. Patients who have an existing pressure ulcer should be removed from the population as they already represent a case of the condition.
Furthermore, it is worth reiterating that a case of pressure ulcer is defined as a resident with an ulcer, not an individual ulcer. A resident with multiple pressure ulcers is counted as one case of the condition. Many of the residents in both of our studies had more than one ulcer.
We believe that residents in the incidence study had risk factors for the development of pressure ulcers similar to those residents in the prevalence study. To lend support to this belief, we re-examined unpublished demographic data on the residents in the prevalence and incidence studies in both LTCFs and found no differences in the primary diagnosis between either study group (see Table 1). We conclude that the populations in both prevalence and incidence studies were equally at risk and that the incidence rates reported for both LTCFs are accurate and not underestimated as suggested by Halfens and Bours.
We concur with Halfens and Bours that incidence rates give a good insight into the quality of preventative care but not necessarily about the quality of treatment. However, we think it is reasonable to conclude that if best practice guidelines for pressure ulcer prevention are implemented in a particular facility, a high probability exists that best practices for treatment are also carried out. The nurses and healthcare aides involved in the care of residents in our studies received similar training in pressure ulcer prevention and skin care management. Similar protocols were in place for pressure ulcer prevention and treatment in both LTCFs. We reject Halfens and Bours insinuation that the difference in prevalence between the two facilities was due to difference in the quality of treatment provided. We stand by our statement that the level of nursing care in both LTCFs was equal and acceptable based on similar low incidence rates. Nonetheless, in our next prevalence study, we plan to record the duration of pressure ulcers, despite the shortcoming of retrospective data, in an attempt to evaluate the effectiveness of treatment.
- Carol M. Davis, Victorian Order of Nurses Toronto York Region Branch Markham, Ontario, Canada and Nigel G. Caseby, Rothbart Pain Management Clinic North York, Ontario, Canada
1. Frantz RA. Measuring prevalence and incidence of pressure ulcers. Advances in Wound Care. 1997;10(1):21-24.