Introduction

Top Abstract Introduction Methods Results Discussion References

There are many superstitions in the world based on religion, old story, legend, fortune telling, and personal experience. Such superstitions influence people's behaviour in various social situations. In medicine, studies have been reported on Friday the 13th, the eight month fetus, and others.^1-4 In Japan, Kaku reported that the superstition called Hinoe-Uma caused a significant increase in the induced abortion rate and a sharp decrease in the birthrate in 1966.^5-9

Belief in Taian-Butsumetsu, a superstition relating to the six day lunar calendar, is common among Japanese people and affects various social events in everyday life. For example, people tend to have wedding ceremonies on Taian (a lucky day), avoiding Butsumetsu. The figure shows the six day cycle of the lunar calendar (Sensho, Tomobiki, Senpu, Butsumetsu, Taian, and Shakku) related to the solar (Gregorian) calendar system, which was adopted in Japan in 1873. Despite this change Taian-Butsumetsu continued to gain popularity.¹⁰

Patients often express the wish to leave hospital on a Taian day by extending, not shortening, their stay. However, if doctors allow patients who do not need to be in hospital to extend their stay they cause an unnecessary rise in the cost of medical care. To show the effects of this behaviour we calculated the mean number of patients discharged on each day of the six day cycle and estimated the costs of any extended hospital stays.

View larger version (37K): [in this window] [in a new window]   Solar and lunar calendars. Sensho is a good day for urgent business. Tomobiki is a bad omen for funerals and Buddhist services, and to hold a funeral on this day is said to be inviting another death. Senpu can be a bad day unless tasks are performed serenely. Butsumetsu means bad luck in all respects. Taian is day of good fortune in all respects, and Shakku brings bad luck in all respects except at noon.

	Methods

Top Abstract Introduction Methods Results Discussion References

We obtained computerised data from the register of patients discharged from Kyoto University Hospital from 1 April 1992 to 31 March 1995. We calculated the mean number, age, and hospital stay of patients discharged on Sensho, Tomobiki, Senpu, Butsumetsu, Taian and Shakku separately, excluding the discharges due to death. Since fewer patients were known to be discharged at the weekend than during the week the numbers of patients discharged on each day of the six day cycle were adjusted for days of the week as follows. Firstly, the total numbers of days for the study period were shown in a 6x7 tablethat is, the six day cycle in columns and seven day cycle in rows. Next, the proportion of each day in the seven day cycle was calculated. The numbers of patients discharged on each day of the seven day cycle in a column of the six day cycle were multiplied by the corresponding proportion of the total days of the seven day cycle. And finally, these numbers were summed for each column of the six day cycle.

We estimated the effect on costs of medical care at the hospital and nationally, based on the following assumptions. Firstly, the influence of superstition is reflected in the difference between the actual and expected numbers of patients discharged on each day of the six day cycle. Secondly, if a doctor grants a patient's wish to leave hospital on Taian, it would be on the nearest Taian day before or after the initially scheduled date of discharge. Thirdly, the mean daily hospital charge is 4200-16 800 yen (£19-£75, $30-$120) as reported for 1997.¹¹ Fourthly, the effect of superstition would be the same all over Japan. And, finally, the total number of patients who leave hospital in Japan is fixed at 13 000 000/ year.¹²

Statistical methods
We calculated the mean and 95% confidence intervals of the number, age, and hospital stay of discharged patients on each day of the six day cycle. Since these data did not exhibit a normal distribution we used the Kruskal-Wallis test followed by the Mann-Whitney test. The differences between days of the six day cycle with respect to sex, age, and hospital stay were examined by the ² test. A P value less than 0.01 was considered significant. Statistical analysis was performed using the SAS software package version 6.12.

	Results

Top Abstract Introduction Methods Results Discussion References

Of the 23 677 patients discharged from the Kyoto University Hospital during the study period, 12 613 (53.3%) were female and 11 064 (46.7%) were male. The mean number of discharged patients was 21.6 a day with the mean age 42.3 years and the mean hospital stay 37.1 days. The mean number, age, and hospital stay of discharged patients were highest on Taian and lowest on Butsumetsu (25.8 v 19.3 patients/day, P<0.0001; 43.9 v 41.4 years, P<0.0001; and 43.1 v 33.3 days, P<0.0001 respectively; table). The Kruskal-Wallis test showed significant difference among the days of this cycle regarding the mean number (P<0.0001), age (P<0.0001), and hospital stay (P<0.0001) of discharged patients.

The difference between actual and expected number of discharged patients on Taian days was about 771 patients over the three years. Thus about 3.3% of discharged patients were estimated to have changed their discharge day because of their belief in Taian. The ratio of the female to male patients on Taian was higher than that on other days (1.19 v 1.13), although this was not significant. The ratio of discharged patients aged 50 years to those <50 years on Taian was significantly higher than that on other days (0.92 v 0.80; P=0.001 by ² test). The ratio of hospital stays 29 days to those <29 days on Taian was significantly higher than that on other days (0.71 v 0.53; P=0.001). The number of women aged 50 or more with hospital stays of at least 29 days who were discharged on Taian was 2.15 times higher than the number discharged on Butsumetsu.

Economic estimation
Given that the mean hospital charge was 12 600 yen a day and that the extra patients extended their hospital stay to the nearest Taian, we estimated that the extra cost to our hospital was 7.4 million yen (£31 000) a year. However, if all the patients shortened their hospital stay to the nearest Taian, although much less likely, the saving would be 12.1 million yen.

	Discussion

Top Abstract Introduction Methods Results Discussion References

We found that significantly more patients were discharged on lucky days and fewer on unlucky days. There were significantly more older patients, especially women, discharged on Taian than on Butsumetsu. Since more older people tend to believe superstitions than young people, this finding further supports the intentional shift of discharge day to Taian by superstitious patients. It would be interesting to study whether the superstition affects date of admission and operation.

Our data do not directly show that superstitious patients extended their stay. However, it is a reasonable inference that they did so because of the significantly longer hospital stay among the discharged patients on Taian as well as our clinical experiences. If the extra patients did extend their hospital stay to the nearest Taian, the cost nationally would be 1400-5320 million yen a year, accounting for 0.005-0.019% of the total national medical costs in 1997. Thus the Taian-Butsumetsu superstition had an important effect on the cost of medical care, although only 3.3% of discharged patients contributed to this phenomenon.

Effect of extended stay
For patients, an extended hospital stay does not necessarily constitute financial burden because the public medical insurance system in Japan covers most of the costs. Even extremely costly care is covered by public medical insurance system once the total cost exceeds a certain level each month. This means that however much their care costs the amount paid by patients is legally fixed at a low level and they do not see the true cost of unnecessarily long hospital stays.

Psychological effect of superstition
An action against superstitious belief could have a negative effect on the health of some patients. In a study on the relation between superstition and health, Scanlon et al reported that there were more hospital admissions as a result of traffic accidents on Friday the 13th (an unlucky day in Western culture) than on other Fridays.² Superstition may have a similar influence on health as the placebo effect, in which patients' health improves after a dummy treatment because of their expectations or belief.^14-16 Furthermore, if patients have a say in when they are discharged from hospital, the "locus of control" is put in their hands to some extent, which may add to the beneficial effect. Strong action against superstition, however, may distress patients and lead to a deterioration in their health.¹⁷ In this context, doctors should pay special attention to patients with somatisation (such as those with depression or anxiety disorders). The relationship between Taian-Butsumetsu belief and health outcome is an intriguing question for prospective study.

Conclusion
Belief in Taian-Butsumetsu certainly influences the decision to leave hospital. The superstition related behaviours of patients cause financial and complex psychosocial problems. Although the costs of unnecessarily extended hospital stay are appreciable, no data exist on the psychological impact of the superstition. Doctors should explain drawbacks of unnecessarily extending hospital stay to the patients while taking into consideration the possible psychological impact of the superstition on health.

	Acknowledgments

This study was presented at the 2nd Asian-Pacific congress of epidemiology, Tokyo, Japan, on 28 January 1998. We thank Mr Gert van Tonder of the Graduate School of Human and Environmental Studies at Kyoto University for his linguistic help and Mr Masahiro Natsume of Medical Informatics Department at Kyoto University (Kyoso Computer Co Ltd ES3 Implementation Department) for data collection.

Contributors: KH initiated the research, conducted analysis of the data and wrote the paper. TF participated in all processes and is guarantor for this paper. AE participated in data collection and discussion. MR and MM participated in the analysis of the data and discussion.

Funding: None

Competing interest: None declared.

	References

Top Abstract Introduction Methods Results Discussion References

1. Scott R, Scott H. Crocodile bites and traditional beliefs in Korogwe District, Tanzania. BMJ 1994; 309: 1691-1692[Abstract/Free Full Text].
2. Scanlon TJ, Luben RN, Scanlon FL, Singleton N. Is Friday the 13th bad for your health? BMJ 1993; 307: 1584-1586.
3. Reiss RE, Ash AD. The eighth-month fetus: classical sources for a modern superstition. Obstet Gynecol 1988; 71: 270-273[Medline].
4. Blacher RS. Clusters of disaster: superstition and the physician. Gen Hosp Psychiatry 1983; 5: 279-284[Medline].
5. Kim YS. Fertility of the Korean populationin Japan influenced by a folk superstition in 1966. J Biosoc Sci 1979; 11: 457-464[Medline].
6. Kaku K. Were girl babies sacrificed to a folk superstition in 1966 in Japan? Ann Hum Biol 1975; 2: 391-393[Medline].
7. Kaku K. Increased induced abortion rate in 1966, an aspect of a Japanese folk superstition. Ann Hum Biol 1975; 2: 111-115[Medline].
8. Kaku K, Matsumoto YS. Influence of a folk superstition on fertility of Japanese in California and Hawaii, 1966. Am J Public Health 1975; 65: 170-174[Free Full Text].
9. Kaku K. Are physicians sympathetic to superstition? A study of Hinoe-Uma. Soc Biol 1972; 19: 60-64[Medline].
10. Uchida M. The calendar's tale for 12 months. Tokyo: Yuzankaku, 1991:106-110.
11. Survey Department of Institute of Social Insurance. Drug tariff. Tokyo: Institute of Social Insurance , 1997.
12. Statistics and Information Department. Patient survey. Tokyo: Ministry of Health and Welfare, 1993:566.
13. Organisation for Economic Cooperation and Development. OECD health system facts and trends1960-1991. Paris: OECD, 1993:215.
14. Oh VM. The placebo effect: can we use it better? BMJ 1994; 309: 69-70[Free Full Text].
15. Dobrilla G, Scarpignato C. Placebo and placebo effect: their impact on the evaluation of drug response in patients. Dig Dis 1994; 12: 368-377[Medline].
16. Kleijnen J, de Craen AJ, van Everdingen J, Krol L. Placebo effect in double-blind clinical trials: a review of interactions with medications. Lancet 1994; 344: 1347-1349[Medline].
17. Edward A, Walker JK, Wayne JK. Psychological factors affecting medical conditions and stress responses. In: Stoudemire A, ed. Human behavior. Philadelphia: Lippincott, 1990:64-83.