Hospital length stay : register-based studies on breast cancer surgery

The present investigation focused on three main questions: what factors can determine length of stay for primary surgery for breast cancer? Can these factors be attributed to the decreasing trend in mean length of stay for primary breast-cancer surgery? And can there be any adverse effect of declining length of stay, such as prolonged sick leave, following breast-cancer surgery? For the four studies of the thesis, routinely collected data was used. In study I the Swedish Hospital Discharge Register was used to describe the trend in mean length of stay of women surgically treated for breast cancer and to investigate the effect of the increasing use of breast-conserving surgery on the decreasing trend in mean length of stay. In study II the effect of stage of disease on length of stay was investigated using the Southern Swedish Tumour Register and the Swedish Hospital Discharge Register. To study the possible adverse effect on prolonged sick leave of decreasing length of stay, data from the National Cancer Register, the Sick Leave Register and the National Hospital Discharge Register were combined (study III). In study IV the effect of social factors on length of stay was investigated using the Swedish Social Medicine Database. In terms of determinant of length of stay, the present results support the hypothesis that individual factors such as age, severity of cancer according to tumour size and social factors influence length of stay. Differences in length of stay according to clinical practice style, such as different types of surgery, were found. There was a difference of about 30 per cent in length of stay between breast-conserving surgery and mastectomy. The results indicate a small effect of change in type of surgery on the mean-length-of-stay trend. In the two studies investigating the effect of changes in surgery on mean length of stay, the calculated effects of the changes in techniques were 14 and 13 per cent, respectively. Cancer stage distribution did not affect the trend. In study III one possible adverse effect of reduced hospital stay was analyzed, namely the number of sick leave days after discharge. A statistical correlation between median length of stay for a particular hospital and number of sick leave days following discharge was established. Shorter median length of stay for a hospital is related to longer sick leave periods following care, indicating a possible adverse effect of shorter length of stay. But long median length of stay for a hospital was also related to longer sick leave period