In this study, we observed a stable incidence of wrist fractures among adults in Norway in the period 2009–2014, with a mean age- and sex-adjusted rate of 244 wrist fractures per 100 000 inhabitants. This is the same incidence as in Sweden and Finland in the same period (5, 6). Unchanged or reduced incidence is also seen in other countries, despite the aging population (2, 6). This trend is assumed to be attributable to a strong focus on prevention of osteoporosis in post-menopausal women through hormone replacement therapy, use of vitamin D and calcium, and also to the fact that they are more physically active (2, 6).
As expected on the basis of previous studies (2, 5), women accounted for around 3/4 of the patient population, with a higher mean age than men. Men have a substantially lower incidence of wrist fractures throughout their lives. The difference in incidence between men and post-menopausal women is assumed to be partly attributable to the difference in the incidence of osteoporosis. The incidence of wrist fractures varies relatively little among the catchment areas in Norway: Førde and Bergen, which had the highest rates, had a 1.3 times higher fracture incidence than Telemark, with the lowest rate.
In a countrywide, population-based historic cohort study of 1 000 hip fractures in Norway in which the data in the Norwegian Patient Registry were validated, high sensitivity for identification of hip fracture was found when the latter was defined using a combination of ICD-10 diagnostic codes and NOMESCO procedures codes (21). A combination of codes was also used in the present study. Thus, the difference in incidence that was identified cannot be explained as due to inaccuracy in the recording of patient data. In principle, treatment of wrist fractures is not sensitive to either preference or supply, since the treatment must be regarded as a necessary health service for those affected. A necessary health service is characterised by a documented effect, that there is little or no disagreement on the treatment and that the benefit exceeds any side effects or negative consequences (22). Nor, then, is the variation in incidence of wrist fractures likely to be due to patients in some areas showing a greater tendency to seek doctors attention after an injury. In the population-based study Cohort Norway, with more than 180 000 participants, it was found that the incidence of forearm fractures in both sexes rose with increasing urbanisation, and that patients in rural areas had higher bone density than patients in urban areas (23). However, this does not explain the differences in incidence observed in this study, since the catchment areas include both urban and rural areas.
Countrywide, the percentage of the total patient population with wrist fractures who received operative treatment was stable at around 28 % during the study period. This is in contrast to what is seen in other countries, where an increasing proportion undergo operative treatment (3). Figures from the Swedish fracture registry for the period 2004–2010 revealed a 40 % increase in operative treatment of wrist fractures concurrently with a fall in incidence (5). Figures from the Finnish Patient Registry showed a 50 % increase in the period 1998–2008 (24). The Swedish and Finnish studies did not specify the percentage of the total patient population with wrist fractures who undergo operative treatment. As with the figures from the Swedish fracture registry, there are small differences in sex and age composition between those who receive conservative treatment for wrist fractures and those who undergo operative treatment (5). An American study of Medicare data shows the same tendency, with the same operation rate as in Norway (17).
We found that there was variation in operation rates across the catchment areas of the various health authorities. Helgeland, with an operation rate of 16 %, and Førde and Vestfold with around 40 %, stood out. Defining observed variation as not justified or undesirable is difficult if the optimal treatment has not been clearly defined, as is the case for many treatment options (25–29). Unjustified variation in the use of health services implies an unequal distribution of health resources (30). In October 2013, the Norwegian Orthopaedic Association published treatment guidelines as decision-making support for treatment choices (18).
Whereas there was no change in the percentage undergoing operative treatment in the period 2009–2014, there was a change in the type of operative procedure chosen. We saw an increase in the use of plates, from 53 % to 81 % (Fig. 1), at the expense of percutaneous pinning and external fixation. Similar findings were made in Sweden, where the proportion for whom plates were used increased from 16 % in 2004 to 70 % in 2010 (5). The choice of volar locking plates rather than pins and external fixation is supported by metaanalyses published in the Norwegian treatment guidelines for wrist fractures (18).
This study showed that in catchment areas with the highest proportion of operative treatment and use of plates, both the operation rate and the rate of treatment with plates were reduced in the last year of the period. Similarly, the operation rate and rate of treatment with plates increased in some catchment areas with an operation rate lower than the national mean. This, too, indicates a change in the situation with respect to indication for surgery, but also a change in preferred surgical method. There is some literature to suggest that undesirable variation can be improved by receiving feedback about one's own practice (26). The orthopaedic surgery community has received feedback of this kind indirectly through the treatment guidelines and the attention they have received. Further changes can be expected in the time ahead.
It is difficult to establish with certainty what a correct treatment level is. The mean treatment rate in Norway is not necessarily the correct level, but it can be assumed that the correct level lies closer to this mean than to the extremes. Therefore it will probably be possible to raise the quality of treatment most by adjusting the levels for patients in the catchment areas with lowest and highest consumption towards the national mean. Such a harmonisation of practice, which we see in the last year of the period 2009–2014, will probably mean higher quality for the patient population as a whole.
This study does not evaluate patients' experience of the treatment they have received. Fifty-eight per cent of all complaints to the Norwegian System of Patient Injury Compensation end in rejections (15, 16). This indicates that there are large differences between patients' expectations of the outcome of their treatment, and the actual outcome. Patients should be involved to a greater extent in the decision concerning treatment, and encouraged to make choices that are more in line with their preferences (31).