When considering the first 20 months, the ICER is $3,830 per QALY. The dance intervention was shown to increase the QOL, which translated into a QALY gain of 0.10 after 20 months. In addition to an increase in QOL, sustained new healthy habits for this target group might prevent future psychiatric illness. The aim of this analysis was not to evaluate which elements of the intervention that was important for the effect.
There is no official level of willingness to pay for a gained QALY in the United States, but $50,000 and $100,000 are often used. In Sweden, an unofficial threshold of $75,000 has been used to guide decisions about subsidized medicine. The ICER of $3,830 per QALY in this study is consequently well below that threshold. These results indicate that the dance intervention was valuable and an efficient use of healthcare resources in relation to what Western countries are willing to pay for a gained QALY.
Healthcare consumption, i.e. number of visits to the school nurse, was an uncertain factor. A test of the factor showed that there was a considerable gap between the number of visits reported by the girls and the number of visits found in the medical records. The school nurses expressed difficulties with the charting of social and mental health issues, due to lack of time, tradition, structure of the journal, and ethical considerations[31, 43]. The self-reported visits were based on a retrospective open question which can be unvalid. Therefore, the ICER was calculated without these, giving an ICER of $6,700. The sensitivity analysis also showed that the ICER was doubled ($7,660) when using 50% of gained QALY and $7,180 when costs were assumed to be 50% higher than calculated, still well below the threshold value.
In this analysis, only savings for the School Health Services were considered. The larger increase in QOL in the intervention group may also have led to savings for other institutions such as primary care, youth centers, and the total Welfare Services in school. Since research has shown associations between internalizing problems in adolescence and in adulthood, it can also be assumed that the larger increase in QOL after the dance intervention could lead to further savings[44–46]. However, the present study was too small to be able to support this notion.
The girls included in the study were obviously exposed to outside influences in addition to the intervention. For example, periods of examinations in school may have increased feelings of stress and thereby affected the girls’ general well-being. Since the intensity of school work varies during the school year, the 12-month follow-up can be considered the most important one.
Strengths and weaknesses
The economic evaluation was performed as a cost-utility analysis from a societal perspective, based on a randomized controlled trial. This method is adopted to find out how you can optimally allocate limited resources and is suitable when the aim is to compare different kinds of interventions. It is also a preferred approach for stakeholders compared with other common models[26, 47].
The HUI3 is a well recognized instrument, published in detail[27, 32, 33, 48, 49], and providing a good estimate of utility values in a community-dwelling, relatively healthy, population. It has been developed by using preferences from a random sample of respondents 16 years of age and older and has also been used to measure health status in several studies on children. However, a weakness is that the validity and reliability of the Swedish version of the instrument have not been tested. We did not agree with the translation in the official Swedish version of the word ‘unhappy’. In the official version, ‘nedstämd’ (similar to lightly depressed) was used, but we chose to use what we believe is a more common expression among youth, ‘olycklig’ (similar to not happy or not glad).
A strength of this analysis is its long-term perspective. There is a great need for studies that establish the sustainability of interventions, to evaluate whether the effects can remain for a long period after the end of the intervention. In this case the 20-month follow-up was approximately a year after the last session of the intervention.
Extrapolation of the trial data was necessary. The last observation carried forward approach was used to handle missing data. This approach has been criticised in recent years for weaknesses in validity and estimations of the result[37, 39, 50] . Other methods were considered but the last observation carried forward approach was chosen since both groups in the study were feeling better after 20 months. Since the participants in the study in general showed an increase in QOL, it can be assumed that the results are more likely to be underestimated than overestimated, which some of the critics claim often is the case.
It cannot be ruled out that differences in QOL at baseline, despite not being significant, may have had an impact on the cost-effectiveness ratio. A lower value at baseline means a higher possible increase. Moreover, it cannot be ruled out that a control group with exactly the same QOL at baseline would have increased their QOL more than the control group in our study. Hence, some caution in interpreting the results is recommended.
Differences in gains of QOL between the intervention and the control group were 0.06 units at the end of the intervention and 0.08 one year later (follow-up at 20 months after start). Hence, a sustainable effect of the intervention seems to be possible, which would in that case give a lower ICER. However, we have not found any research which indicates for how long time the effect may remain. We don’t think it is likely that the effect will end the day after the last follow-up. On the other hand, assumptions of very long-lasting effects may be an overestimation.
In the way the intervention was organized in this study, there were no costs for the participants. The dance hall was close enough to the participants’ schools to not require any travelling costs. The girls in the intervention reported high enjoyment of the dance sessions. This indicates that there might not have been any sacrifice, such as loss of enjoyment compared to an alternative activity, to participate in the dance sessions.
Results in relation to previous research
No economic analyses were found of dance as an intervention to prevent or treat internalizing problems. This analysis therefore seems to be filling a gap of knowledge.
The scarceness of evidence-based interventions for adolescent girls with internalizing problems may be due to the lack of studies as well as organizational complexity, with many institutions involved in prevention and care. However, the patient group is in great need of effective interventions.
Further research required
The present study evaluated the cost-effectiveness of an intervention with dance for adolescent girls with internalizing problems. Intervention studies and economic analyses of interventions for this target group is an unexplored research area, in great need of further studies. Specifically for this study, there is also a need to find out what the most important components of the dance interventions were.
The results of this research can support comparisons with other approaches to prevent or treat mental problems as well as with other modalities. The research also highlights the need for long-term research to establish the sustainability of this type of interventions.
The present study also revealed that there is a lack of research on the school health services in Sweden.