It is well known that Helicobacter pylori (H. pylori) infection is the major cause of gastric ulcers (GU) and eradication of the infection can reduce the occurrence and recurrence of GU. Many different therapeutic strategies performed might have caused some confusion for therapists, hence causing some disadvantages for cost-effectiveness to public health care. To standardize the therapy for GU patients in Japan, an evidence-based guideline (GL) with the provision of an explanatory flow chart (Fig. 1) was first published in 2003 (1st version) (1) and then in 2007 (2nd version)(2) with a minor revision under the support of the Japanese Ministry of Health, Labor and Welfare (GUGLJ).

Fig. 1. Flow chart of treatments in the EBM-based Guideline for Gastric Ulcer in Japan.

It is believed that clinical practice with GLs can improve the quality, appropriateness and cost-effectiveness of health care, and can also serve as valuable education tools. Among 279 GL published from 1985 to 1997, however, the mean adherence to standards by each GL was low, estimated at 43.1% according to the standards for GL development established by Shaneyfelt et al. in 1999 (3). On the other hand, adherence of GUGLJ to these standards is high, estimated at 80% (1). Theoretically, the degree of doctor’s knowledge on the GUGLJ and their respective specialties may relate to the degree of their adherence to the GUGLJ for the treatment of the GU patients. A stronger adherence of doctors in charge indicate a stronger adherence to treatment performed for GU patients, which may have resulted in more effective treatment of the GU. Effectiveness means managed and improved control of GU with less expense of the therapy performed. Assessment and eradication of H. pylori is important to control gastric cancer development as well, which can follow the H. pylori induced chronic gastritis (4). The present study herein has been designed to perform a multihospital analysis to assess the relationship of a doctor’s knowledge and specialty for the GUGLJ and the medical cost utilized for the complete treatment of a GU patient, and is referred to as “Japan assessment study on the evidence-based guideline for gastric ulcer with special reference to its outcome: EGGU.” To date, no other papers surveying the nationwide cost-effectiveness of GUGLJ-based therapy of GU patients have been published in Japan.


MATERIALS AND METHODS
Gastric ulcer patients and doctors in charge

The present study was a retrospective cohort analysis. GU patients with active ulcers diagnosed by an endoscopic examination between September 2004 and April 2005, and doctors in charge that treated the patients at NHO hospitals were enrolled through an internet registration on the EGGU web site between May 2006 and March 2007. Patients were selected to be 20 years old on April 1, 2006, or 18 years old at the time of GU diagnosed.

Patients were examined for their subjective symptoms relating to the ulcer healing or recurrence when they were enrolled in the present study after the treatment of GU by inquiring about their health conditions via mail between May 2006 and March 2007. Health status assessment was performed using SF36, version 2 (5). Those doctors in charge enrolled in the study were surveyed by the EGGU researcher at each hospital by their specialty, whether gastrointestinal (GI) experts, non-GI physicians or other specialists. Their knowledge regarding the GUGLJ was divided into 3 groups based on by the answer to the inquiries about the GUGLJ: knew it very well, just yes or no. Medical records of each patient were examined to discover the endoscopic findings of the ulcer healing or recurrence, and the medical treatments performed. Adherence scores to the GUGLJ were adapted to each category of performance for the therapy of GU. The total adherence scores were calculated by counting the number of performances and their adherence scores. The present study was undertaken with permission from the ethical committee of the NHO in Japan. All doctors and patients involved with the study had signed their informed consents.

Flow chart described in the guideline

Active ulcers are initially checked to determine if they bleed or not. After the bleeding stops or no bleeding is recognized, assessment for non-steroidal anti-inflammatory drugs (NSAID) intake and H. pylori infection is undertaken. If H. pylori infection is positively identified, and its eradication is performed successfully, no maintenance therapy is requested following the eradication. This flow chart can clearly inform the doctors in charge of the treatment strategy of the GU patient (Fig. 1).

Guideline adherence scores

In order to calculate the GUGLJ adherence scores, five items shown in Table 1 were scored. They were 1) NSAID; 2) H. pylori infection; 3) ulcer healing treatment; 4) maintenance therapy; and 5) therapy for patients with no H. pylori infection. “Comments” in Table 1 indicated the different activities of the doctors in charge or therapies performed; and “Therapies” indicated the drugs prescribed to the GU patients. For example, the presence of description about NSAID intake or the assessment for H. pylori infection by the doctor in charge in the medical records was scored 5. On the other hand, no description in the medical records about these two items was scored 0. These scores were determined by the significance for the therapy, which is reflected in the flow chart of the GUGLJ. Finally, the total GUGLJ adherence scores counted were divided into 4 groups: category 1 with a total score of 5, category 2 with a total score of 6 and 12, category 3 with a total score of 13 and 20, and category 4 with a total score of 21; the larger category number indicates greater adherences to the GUGLJ. These categories were drawn from the retrospective assessments of the medical record of each GU patient, not from the subjective memory of the doctors in charge. Compliance of the GU patients was not analyzed because it was subjective and difficult to estimate. Subjective data were limited to the health status assessment of GU patients using SF36 version 2. Since the adherence scores and categories were counted utilizing only the objective findings described in the medical record of GU patients, they should be highly reliable.

Table 1. Guideline adherence scores.

Abbreviations: NSAID, non-steroidal anti-inflammatory drugs; PPI, proton pump inhibitors; H2B, H2 blocker (histamine H2 receptor antagonists); MP, mucosal protectants; MTx, maintenance therapy

Health status assessment

For the health status assessment of GU patients, SF36 version 2 (5) was applied. This is a program accepted worldwide to assess the health-related quality of life (QOL) of patients. SF36 is a popular tool to compare the patient’s physical and mental condition among different diseases. It includes eight subscales: physical functioning (PF), role physical (RP), bodily pain (BP), social functioning (SF), general health perceptions (GH), vitality (VT), role emotional (RE) and mental health (NH), and two summary scores: physical component summary (PCS) and mental component summary (MCS). Letters of inquiry were delivered to each patient and the answers returned to the EGGU office by mail by patient’s own voluntary participation.

Medical cost utilized for the treatment of gastric ulcer

The relationship between the adherence scores to GUGLJ and the medical cost consumed during the following nine months after the diagnosis of GU was analyzed. All data described in the medical records of GU patients that consulted the NHO hospitals enrolled were analyzed. If the GU patients consulted hospitals other than an NHO hospital, or they consulted the NHO hospitals as result of other conditions, those data were not included in the analysis. When the period of prescription for GU was recorded longer than nine months, the data were excluded from the analysis. Accounting of the medical cost expended was done based on the drugs prescribed, the endoscopic examinations performed, and the medical cost of outpatient clinic and inpatient hospitalization. These accounts were basically done with respect to the National Health Insurance points of Japan. The medical costs of drugs prescribed were accounted for by counting the volume of drugs per day and the duration of drugs prescribed. Generic or non-generic brands of the drugs prescribed were also included in the accounting process. The medical costs of the outpatient clinics were accounted for by counting the number of consultations, and the fees of consultation and prescription with respect to health care insurance. We found that the average duration of hospitalization for GU/duodenal ulcer (DU) was 22.5 days. The National Health Insurance points of GU/DU have been published in the data of the Diagnosis Procedure Combination (DPC) system of the Japanese Ministry of Health, Labor and Welfare. The medical cost consumed for GU/DU therapy in a one-day inpatient hospitalization was accounted for by dividing the National Health Insurance points of GU/DU by 22.5. Then, the total costs of inpatient hospitalization were assessed with that of one-day inpatient hospitalization and the duration of inpatient hospitalization recorded. Finally, the cost of one endoscopic examination, one-day consultation for GU, and one-day inpatient hospitalization for GU were estimated at 11,400; 1,380; and 26,410 Japanese Yen, respectively.

Statistical analysis

The base-line characteristics of the patients and doctors enrolled were compared by the Student’s t-test. Recurrence rates of GU, the medical performances and the GUGLJ adherence scores, the description of whether the patient was a NSAID taker or not, and the assessment of H. pylori infection were compared by the chi-square test or Student’s t-test. Because the QOL-scores and the medical costs consumed are influenced by gender, age and complications of the GU patients, and whether the patients were NSAID taker or not, the associations between the GUGLJ adherence score and PCS/MCS or the medical cost consumed were investigated by adjusting these factors by a multiple regression analysis using software of JMP8 (SAS Institute Inc., Cary, NC, USA) or STATA11 (StataCorp., Texas, USA), respectively. Different medical costs expended of category 1 to 4 were investigated by the Kruskal-Wallis test. Significance was indicated by a P value of less than 0.05.


RESULTS
Patients

Initially, 942 GU patients were enrolled but seven patients were dropped out during the follow-up period. Therefore, a total of 935 patients (572 male, 363 female, average age =63.9 years) were enrolled. Out of these GU patients, 283 (30.3%) had a past history of GU and 554 (59.3%) had complications; diabetes in 100, hypertension in 176, heart disease in 107, vascular disease in 107, liver disease in 79 and others in 302. For the GU patients, description whether the patient was an NSAID-taker or not was found in 847 (90.6%) of their medical records, and 162 (17.3%) were NSAID-takers. H. pylori infection assessment was performed in 622 (66.5%). Of these GU patients, 75.6% were H. pylori positive and 24.4% were negative. Assessment of eradication was performed in 81.3% of H. pylori-positive GU patients, and 86.8% of them succeeded in total eradication.

Doctors in charge of the gastric ulcer patients

In total, 270 doctors (232 male, 38 female, average age =40.9) from 62 NHO hospitals were enrolled. There were 173 (64.1%) GI experts, and 70 (25.9%) non-GI physicians. The doctors in other fields numbered 27 (10.0%). Regarding inquiries about the GUGLJ, 79 (29.3%) doctors in charge answered “yes very well.”, 128 (47.4%) answered just “yes.” and 63 (23.3%) answered “no.”

Recurrence of gastric ulcer with respect to NSAID status and eradication of Helicobacter pylori

NSAID-takers tended to show higher GU recurrence rates than NSAID non-takers, but there was no statistical significance between them (Table 2). On the contrary, the eradication may contribute to the reduction of the GU recurrence in the H. pylori-positive GU patients. Among H. pylori-positive patients, the GU recurrence rates were significantly lower in patients with successful eradication than those with failure eradication either NSAID takers or not (Table 3).

Table 2. Relationships among NSAIDs intake and recurrence in the Helicobacter pylori negative gastric ulcer patients.

Abbreviations: NSAID, non-steroidal anti-inflammatory drugs; PPI, proton pump inhibitors

Table 3. Relationships among NSAID intake, eradication and recurrence in Helicobactor pylori positive gastric ulcer patients.

aP=0.003; bP=0.03; cP=0.04 Abbreviations: NSAID, non-steroidal anti-inflammatory drugs; PPI, proton pump inhibitors

Specialty and performance of doctors in charge of the gasatric ulcer patients

Table 4 shows that a description of a GU patient being an NSAID-taker or not was related significantly (P<0.001) to the specialty of the doctors in charge. On the other hand, the assessment of H. pylori infection was carried out most frequently (P=0.002) by the doctors in charge who knew the GUGLJ very well. There were no statistically significant relationships between these performances and the GU recurrence (data not shown).

Table 4. Relationships between specialities of doctors and their performance among 927 patients treated.

Abbreviations: NSAID, non-steroidal anti-inflammatory drugs; H, Helicobacter; GI, gastrointestinal experts; non-GI, non-GI physicians; Does the doctor know about the guideline (GL)? Yv, Yes. I know it very well.; Y, Yes. I know it.; N, No. I do not know it.

Therapeutic drugs prescribed

Proton pump inhibitors (PPI) were most frequently prescribed in 30.4% of the 935 GU patients. Among 162 NSAID-takers, 51.2% quitted NSAID intake after the diagnosis of GU. No maintenance therapy was undertaken in 24.4% of 250 H. pylori-positive GU patients with successful eradication. Among them, histamine H2 receptor antagonists (H2RA) were prescribed most frequently. On the other hand, among H. pylori-positive GU patients with failure eradication and H. pylori-positive GU patients with no assessment of the eradication performed, no maintenance therapy was undertaken less frequently, and a PPI was prescribed in double frequency as compared with those for the former 250 GU patients. Among H. pylori-negative patients, a PPI was prescribed most frequently (Table 5).

Table 5. Therapies for gastric ulcer patients

Abbreviations: NSAID, non-steroidal anti-inflammatory drugs; PPI, proton pump inhibitors; H2RA, histamine H2 receptor antagonists; MP, mucosal protectants; MTx, maintenance therapy

Guideline adherence scores

In total, 927 GU patients were analyzed for the GUGLJ adherence scores. The scores showed a significant relationship to the specialty (P<0.05), and the knowledge (P<0.001) of the doctors in charge (Table 6). There were no statistically significant relationships identified between these categories and the GU recurrence (data not shown).

Table 6. Relationships between guideline adherence scores and number of patients treated

Abbreviations: GI, gastrointestinal experts; non-GI, non-GI physicians; Av, average of Category 1, Scores 5; 2, Scores 6 and 12; 3, Scores 13 and 20; 4, Scores 21; Does doctor know about the guideline (GL)? Yv, Yes. I know it very well.; Y, Yes. I know it.; N, No. I do not know it.

Quality of life of the gastric ulcer patients

To assess the health-related quality of life (QOL) of the GU patients, the SF-36 version 2 was applied. Results indicated that the highest point was scored in the category 4 in all of the subscales (Fig. 2). These subscale scores tended to be higher in male patients, less than 65 of age, no GU recurrence and NSAID non-takers. Then, the relationship between the subscale scores and the GUGLJ adherence scores were examined by a multiple regression analysis with respect to the gender, age and complications of the GU patients, and whether they were NSAID-takers or not. The subscales showed a significant relationship (P<0.01) to the PCS, but not to the MCS of SF-36 (data not shown).

Fig. 2. Guideline adherence scores and the subscales of SF-36. Category 1, Scores 5; 2, Scores 6 and 12; 3, Scores 13 and 20; 4, Scores 21; PF, physical functioning; RP, role physical; BP, bodily pain; SF, social functioning; GH, general health perceptions; VT, vitality; RE, role emotional; NH, mental health.

Direct medical cost expended for the treatment of gastric ulcers

There was a significant difference (P<0.001) in the medical cost consumed among the four categories of the adherence to the GUGLJ scores (Table 7). The total medical cost for these performances in category 4 was calculated as 237,467 Japanese yen, which was 67% of that in category 1, and the least expensive among the four categories with statistical significance (P<0.001) (Fig. 3). The complications of the GU patients, whether they were NSAID takers or not, and the GUGLJ adherence score were associated with total medical cost (Table 8).

Table 7. Relationships between guideline adherence scores and medical costs consumed for treatment of gastric ulcer.

aP<0.001, bP=0.037, cP<0.001, dP=0.013; Abbreviations: Category 1, Scores 5; 2, Scores 6 and 12; 3, Scores 13 and 20; 4, Scores 21;

Fig. 3. Guideline adherence scores and the direct medical cost consumed for the GU therapy. Category 1, Scores 5; 2, Scores 6 and 12; 3, Scores 13 and 20; 4, Scores 21.

Table 8. Multiple regression analysis for total medical cost expended.

Total medical cost was log transformed. Adjusted R squared=0.0834 Age and gender were excluded for the final model because of no statistical significance in the multivariate model. Abbreviations: Category 1, Scores 5; 2, Scores 6 and 12; 3, Scores 13 and 20; 4, Scores 21;

DISCUSSION
H. pylori is an important factor in the pathogenesis of gastro-intestinal disorders.(6, 7). In the past, H. pylori infection has been detected in a range of 56 to 96% of GU patients. This difference may be reflecting the different use of the NSAID among the GU patients (8). In the present study, H. pylori infection was detected in 75.6% and NSAID intake was found in 17.3% of GU patients. H. pylori and NSAID are known to correlate with the GU recurrence. H. pylori infection can increase the risk of GU by an odds ratio of 3.2 (9), and the risk of GU in H. pylori-infected NSAID takers was estimated at 61.1 times higher compared with H. pylori-negative NSAID non-takers (10). Concerning the use of NSAIDs is the basic issue for GU treatment. The GUGLJ recommends the doctors in charge to perform GU therapy after quitting NSAID intake, if possible. In practice in the present study, 90.6% of GU patients were examined for NSAID intake. The doctors in the other fields than the physicians recorded it less frequently. 51.2% of the NSAID-takers quitted the use of NSAID and 79.0% of them received the PPI treatment. PPI treatment is the first choice for the GU patients with no H. pylori- infection or failure eradication, and H2RA is the second choice in case of failure eradication according to the GUGLJ. In the present study, PPI and H2RA were prescribed most frequently for the GU patients with no H. pylori-infection and failure eradication, respectively. H. pylori assessment was performed in 66.5% of GU patients. Doctors who knew the GUGLJ very well did this assessment most frequently. These different specialties or knowledge of doctors in charge may relate to the different GU treatment or the low performance rate of H. pylori assessment. In fact, the GI-experts and the doctors in charge that knew the GUGLJ very well treated the most GU patients in category 4.

In case of no H. pylori infection, GU recurred in 12.1% of patients, which was confirmed with an endoscopic examination. When the failure eradication occurred among the H. pylori-positive GU patients, the endoscopic GU recurrence rates increased significantly up to 50% of NSAID-takers and 29.2% of NSAID non-takers. Therefore, the eradication is important to achieve the effective treatment of GU. The GUGLJ recommended several ways of eradiation and expected that the successful eradication to be performed. However, it does not exclude the failure eradication itself, since the latter still occurs in substantial numbers even if the ideal eradication has been undertaken. In the present study, no scores were adapted to the failure eradication.

Prescribed aids of GU treatment are to relieve the pain, to promote ulcer healing, and to prevent ulcer recurrence. The eradication therapy could heal GU effectively except for the pain relief and reduce the GU recurrence rate after one year compared to the therapy with acid blockers (11). It was reported that the maintenance therapy with acid blockers and mucosal protectants was insufficient to prevent GU recurrence since 24.6% of GU recurred one year after this combination therapy through the primary care and maintenance therapy of GU in Japan. Labenz et al. reported no necessity of the maintenance therapy after the successful eradication of H. pylori (12). However, some doctors in charge of GU patients still use acid blockers and mucosal protectants to protect gastric mucosa since chronic gastritis can continue even after the H. pylori are eradicated. This maintenance therapy would increase the cost of therapy unnecessarily. In the GUGLJ, the maintenance therapy is recommended to be performed only for the GU patients who are not subjects of eradication therapy. Considering the treatment under the limited budgets of health care insurance, the aspect of cost-effectiveness is an essential issue for the overall treatment. However, in the present study, H. pylori infection was assessed for 66.5% of the GU patients, and only 26.7% succeeded in H. pylori eradication. Then, only 24.4% had no maintenance therapy following the eradication while 60.8% received H2RA. These results may indicate the low coverage of GUGLJ among the doctors in charge for GU patients in Japan during the period of the present study performed. Indeed, there were 712 GI experts but only 325 knew GUGLJ well in the present study.

When estimating the cost-effectiveness of medical treatment, it is important to be clear on what the effectiveness is intended for. Basically, the medical cost expended must include not only the direct performance on the patients but also the indirect performance such as loss of social activity relating to the hospitalization of the patients in the hospital. In the present study, however, the economic estimation could cover the direct performances for the GU patients who consulted the NHO hospitals on a regular basis and enrolled within the following nine months after the diagnosis of GU. As a result, the different categories of GU therapy could not influence on the GU recurrence, but may influence on the medical cost consumed.

The different medical costs of categories 1 to 4 were attributed mainly to the different cost of drugs prescribed, and the fees of inpatient hospitalization, the latter of which would reflect the duration of hospitalization. As mentioned before, the GU patients in category 4 complained less PCS of QOL than the other GU patients. This would mean those GU patients in category 4 had fewer complications or complaints relating to the shorter time of GU healing, or both. So, it may be possible to consider that they had the least consumption of drugs prescribed, and the smallest fees of the inpatient hospitalization.

GU therapy may be requested if it recurs even five years after the diagnosis of GU. Indirect influence of the GU patients on the economics would vary according to their social activities. In the present study, the adherence scores of the GU patients were not assessed. The medical costs were evaluated by accounting for the direct performance within limited periods. Therefore, the data obtained in the present study could reflect only the adherence of medical professionals, and then some parts of the total cost expended. However, the results obtained in the present study could reflect the practical medicine performed for the GU therapy in Japan, as evaluated from the view points of medical professionals. The present study could conclude that the therapy for GU along with the GUGLJ is essential to perform a cost-effective medicine and it is most effectively performed by the GI experts or the doctors who understand very well the GUGLJ.

Acknowledgements: The present study named EGGU was supported financially by National Hospital Organization of Japan.

Conflict of interests: None declared.


REFERENCES

1. Research group for assesment and adaptation of guideline for gastric ulcer. Evidence-based guideline for gastric ulcer treatment (in Japanese). Tokyo, Jiho K.K., 2003.
2. Research group for assesment and adaptation of guideline for gastric ulcer. Evidence-based guideline for gastric ulcer treatment (in Japanese). 2nd Ed. Tokyo, Jiho K.K., 2007.
3. Shaneyfelt TM, Mayo-Smith MF, Rothwangl J. Are guidelines following guidelines? The methodological quality of clinical practice guidelines in the peer-reviewed medical literature. JAMA 1999; 281: 1900-1905.
4. Uemura N, Okamoto S, Taniyama K, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001; 345: 784-789.
5. Fukuhara S, Suzukamo Y. Manual of SF-36v2 Japanese version (in Japanese). Kyoto, Institute for Health Outcomes & Process Evaluation Research, 2004.
6. Karczewska E, Wojtas I, Sito E, et al. Assessment of co-existence of Helicobacter pylori and Candida fungi in diseases of the upper gastrointestinal tract. J Physiol Pharmacol 2009; 60(Suppl 6): 33-39.
7. Loster BW, Czesnikiewicz-Guzik M, Bielanski W, et al. Prevalence and characterization of Helicobacter pylori (H. pylori) infection and colonization in dentists. J Physiol Pharmacol 2009; 60(Suppl 8): 13-18.
8. Sugiyama T. Gastric ulcer. In: Helicobacter pylori and the Eradication Therapy (in Japanese). Terano A, Takahashi S, (eds.) Tokyo, Nankodo, 2003, pp. 31-36.
9. Nomura A, Stemmermann GN, Chyou PH, Perez-perez GI, Blaser MJ. Helicobacter pylori infection and the risk for duodenal and gastric ulceration. Ann Intern Med 1994; 120: 977-981.
10. Huang JQ, Sridhar S, Hunt RH. Role of Helicobacter pylori infection and non-steroidal anti-inflammatory drugs in peptic-ulcer disease: a meta-analysis. Lancet 2002; 359: 14-22.
11. Sung JJ, Chung SC, Ling TK, et al. Antibacterial treatment of gastric ulcers associated with Helicobacter pylori. N Engl J Med 1995; 332: 139-142.
12. Labenz J, Leverkus F, Borsch G. Omeprazole plus amoxicillin for cure of Helicobacter pylori infection. Factors influencing the treatment success. Scand J Gastroenterol 1994; 29: 1070-1075.