The increasing population of elderly people with high medical treatment costs throughout Japan is leading to an escalating financial burden.1 Pressure ulcers (PUs) are one of the most striking conditions affecting medical costs and quality of life among patients, specifically bed-ridden patients.2 In October 2002 a new governmental regulation system related to management of pressure ulcers was introduced in Japan. The Japan Ministry of Health, Labour, and Welfare implemented this regulation in all hospitals covered by the National Medical Insurance system. The regulation included a reduced reimbursement of 50 yen (about 48 cents) per patient for each hospitalized day unless the following 3 requirements were met: 1) establishment of a pressure ulcer specialist team including both physicians and nurses, 2) introduction of the standardized clinical protocol for management of pressure ulcers for each patient starting from the time of admission according to their activities of daily living, and 3) provision of adequate bedding support according to the patients’ medical and nursing requirements. This was the first attempt to reduce the medical expenditures using a penalty system and helped raise awareness among medical professionals concerning the management of pressure ulcers. Whether this system actually reduced PU prevalence was not determined.
Surveillance of pressure ulcer prevalence has been conducted by the National Pressure Ulcer Advisory Panel and the European Pressure Ulcer Advisory Panel in the United States and Europe, respectively.3,4 In Japan a study of this proportion has yet to be performed. This is partly due to the lack of social awareness regarding pressure ulcers.5 This particular factor may contribute to the mistaken perception that the PU prevalence was notably low in Japanese hospitals, seeing that only relatively small studies including a limited number of patients at risk for pressure ulcers have been conducted in Japan.
The purpose of this study was to evaluate the governmental regulation system by investigating the PU prevalence using retrospective cohort methods prior to and 1 year after implementation of this regulation. The relationship between the change in prevalence relative to hospital structure and system was also investigated.

Methods
Study hospitals. Five thousand hospitals were identified from all 9230 hospitals in Japan. Hospitals with 300 or more beds were contacted. Hospitals with less than 300 beds were randomly selected. Questionnaires were mailed to each hospital; the percentage of respondents was 51.6.
Study methods. The PU prevalence was calculated as the existing cases of pressure ulcer patients divided by the number of all inpatients during 4 specified time periods; 1 month before the introduction of regulation, immediately after the introduction, 6 months later, and 1 year later. The PU prevalence was calculated in 3 aspects: the PU prevalence before admission, after admission, and total number of pressure ulcers. In Japan, the patients receive initial skin assessment at the time of admission. A pressure ulcer detected at the initial assessment was defined as a pressure ulcer that occurred before admission.
The hospital characteristics that were investigated included the type of hospital, number of patients, number of bed-bound patients, length of stay, and hospital occupancy rate. The condition of bedding support distribution including the type of mattress, thickness, increased bedding support rate, and sufficiency level were also investigated before and 1 year after the introduction of regulation. The increased bedding support rate was calculated as the number of support surfaces 1 year after the introduction divided by the number before introduction. The sufficiency level was defined using the formula: sufficiency level (%) = number of support surface/number of bedridden patients x 100.

Data Analysis
The present study investigates the number of patients with pressure ulcers that developed after admission, before admission, and the total number of ulcers. However, some hospitals were excluded in the calculation of PU prevalence due to the lack of precise responses, such as inconsistencies in the sum at each PU stage and the total number of PUs, a deficient number of inpatients, or lack of hospital description. Due to this manipulation, the total number of PUs did not coincide with the sum of PUs before and after admission.
A mixed-effect model (PROC MIXED with contrast statement) was used to determine the effect of implementing governmental regulation for the differences between each measurement period. For analysis investigating the related factors for improvement of the PU prevalence, the hospitals were divided into 2 groups: improved group and worsened group (according to PU prevalence). The improved group was defined as a hospital showing a reduction in the prevalence between immediately after and 1 year after the introduction of regulation. We also regarded the hospitals with prevalence of 1.0% or less in both immediately and 1 year after the introduction as belonging to the improved group, since this low prevalence could not be further reduced. The worsened group was defined as hospitals that showed an increased prevalence that did not meet the criteria for the improved groups. Significant covariates were entered into a multivariate stepwise logistic regression model to identify those that contain independent information. The threshold values for entry into and removal from the model were 5% and 10%, respectively. A logistic-regression model was used to investigate the independent factors affecting the improvement in PU prevalence.
P < .05 was considered statistically significant. All statistical analyses were performed using Statistical Analysis System Ver. 9.1 (SAS Institute Inc., Cary, NC).

Results

Hospital characteristics. From 5000 eligible hospitals, 2582 hospitals (51.6%) responded. The characteristics of these hospitals are shown in Table 1. As some hospitals provided incomplete responses, the total number of hospitals in tables sometimes differed. Types of hospitals in Japan included university hospitals, acute care, acute and long-term (mixed care), long-term care, psychiatric, and other types. The mean length of stay was 163.7 days. The shortest length was in university hospitals (22 days) and the longest length was in psychiatric hospitals (521.8 days). The mean hospital occupancy rate was 86.8%; the highest was in long-term care hospitals (94.2%).

Prevalence. Overall PU prevalence 1 month before the introduction of regulation was 4.26%. The total PU prevalence gradually decreased to 4.18% immediately after introduction of the regulation, and to 3.95% and 3.64% 6 months and 1 year later, respectively. The prevalence observed at 1 year after the introduction was significantly lower compared with the pre-introduction (P = 0.039, [Table 2]). The changes in PU prevalence after admission were most prominent with a prevalence of 2.77% before the regulation, which gradually decreased to 2.65, 2.56, and 2.31% immediately after, at 6 months and 1 year, respectively, with a significant decrease at 1 year after the regulation compared with before (P = 0.039, [Table 3]). Although a decrease was also observed in PUs before admission, the differences were not statistically significant (Table 4).
Long-term care hospitals showed the highest PU prevalence before and after admission and in total PUs.

Severity. The distribution of severity is summarized in Table 5. The overall proportion of Stage III and IV pressure ulcers was 23.9 and 10.9%, respectively, for PU prevalence after admission but before introduction of regulation. One year after the introduction, the proportion of Stage III and Stage IV ulcers gradually decreased to 18.8% and 8.1%, respectively. Stage II pressure ulcers were the predominant type of PU found in the study.
Pressure ulcer specialist team. Ninety-nine percent of hospitals established a team of pressure ulcer specialists, which indicated that 0.6% of hospitals did not meet the criteria for the PU regulation. This team included physicians, general nurses, wound, ostomy, and continence nurses or enterostomal therapist nurses, pharmacists, dietitians, and hospital administrators. In-service education was conducted in 90.9% of hospitals that had established a specialist team. The multidisciplinary team round was conducted in 73.2% of the hospitals, most of which conducted once or more times per week (46.6%). Fifty-eight percent of the multidisciplinary team round targeted all of the patients who had pressure ulcers.

Distribution of support surfaces. The quantity of each support surface type is shown in Table 6. The alternating air mattress and visco-elastic foam mattress represented a significant share of the support surfaces. The increasing rate of each type of support surface is summarized in Table 7. The sufficiency of support surfaces is shown in Table 8. University hospitals showed the highest sufficiency of 47.5% before introduction and 79.9% after introduction.

Related factors for improved PU prevalence. The hospitals were divided into 2 groups, improved group and worsened group according to the PU prevalence. Comparisons between the 2 groups utilizing single regression analysis revealed that average length of hospital stay and sufficient distribution of air mattresses for bed-bound patients were the key characteristics that were significantly different. For support surfaces, the hospitals were divided into two groups, increased and decreased group, based on the changes in support surface sufficiency before and after the introduction of the regulation. Hospitals were categorized as the increased group if the sufficiency of mattresses increased; otherwise they were put into the decreased group. To investigate the factors that independently affected the improved PU prevalence, a logistic-regression analysis was conducted and revealed that only a sufficient distribution of air mattresses was associated with improvement (odds ratio = 2.259, 95% confidence intervals; 1.091–4.679, P = 0.028, [Table 9]).

Discussion
This surveillance is the first to describe the detailed change in PU prevalence after the implementation of governmental regulation on the management of PUs in Japan. A significant reduction in PU prevalence was observed after the introduction of a reimbursement reduction penalty system. The study also identified sufficient distribution of air mattresses as the factor that most affected the reduction in PU prevalence. The unique point of the governmental regulation was that this system focused on the structure of PU management, not on outcomes or processes.6 Several studies have reported the effectiveness of such regulation on outcomes, such as PU incidence.7 In the US, the Omnibus Budget Reconciliation Act of 1987 was implemented for the process of care in nursing homes; however, there was little reduction in PU prevalence.8 There is little evidence to support the effectiveness of regulating the structure of a management system to reduce PU prevalence.9 The Japanese government introduced this system because of a medical payment system; most hospitals in Japan are covered with a fee-for-service system, where greater PU prevalence will increase a facility’s income. Only some national and university hospitals in Japan use the Diagnosis Procedure Combination system similar to the Diagnosis Related Groups/Prospective Payment System used in US and European countries. Even if the regulation was put on the incidence of pressure ulcers, the hospitals covered with the fee-for-service system may not suffer significant losses as they can be returned the actual fee of PU care. These data, for the first time, demonstrated that regulation on the hospital structure had a beneficial effect on reducing PU prevalence. The authors’ next target is to identify the effect of reduced PU prevalence on overall medical cost savings.
Prevalence and severity. The PU prevalence in Japan is much lower than that of other countries. Many investigators reported the PU prevalence ranged from less than 10% to more than 20% depending on the setting or population investigated.10–12 Although Japan’s prevalence was low, the severity was high. This study cannot clarify the causative factors of this high severity, but physical characteristics may contribute to it: low body weight and extreme bony prominences.
The present findings demonstrated that PU prevalence after admission was significantly decreased after the introduction of regulation; on the other hand, the PU prevalence before admission was not influenced by regulation. This may indicate a high incidence of PUs occurring in the home-care setting; therefore, the next target of PU management should focus on this population.
Related factors. The only factor related to the improved PU prevalence was the sufficient distribution of air mattresses. In Japan, many patients at risk for PU are bed-ridden and have extreme bone prominences, which require sufficient pressure distribution to achieve adequate management.13 The number of visco-elastic foam mattresses or other measures increased; however, they did not independently affect the improved PU prevalence. The possible reason of no independent effect of sufficient distribution of visco-elastic foam mattress or other measures is that they were used not only for pressure ulcer prevention but also other medical requirements. The change in the number of these measures did not represent the improvement of quality of pressure preventive care. The government regulation was focused not only on the bedding support surface system, but also human resources and a standardized care protocol. The other two factors, however, may not affect outcomes within a 1-year period and would need to be determined in future studies.
Limitations. This study has some limitations. Since the present surveillance was conducted 1 year after the implementation of governmental regulation, the real impact on medical costs remains unclear. Additionally, this regulation was conducted in accordance with a government-regulated penalty system, which made it impossible to have a control group. Therefore, a historical control was used and made the strength of evidence relatively low.

Conclusion
The implemented governmental regulation reduced the prevalence of pressure ulcers in Japan. The present study demonstrated modifying the structure of hospitals could influence the quality of outcomes in hospital care.