Conservative or Operative Therapy in Patients With a Fragility Fracture of the Pelvis

Background. Fragility fractures of the pelvis are increasing in incidence. These are osteoporotic fractures, who result from a minor trauma. Typically, the weak osteoporotic bone is fractured, an the ligaments remain intact resulting in an undisplaced or minimally displaced pelvic ring. Fragility fractures of the pelvis can be divided in type I, II, III and IV. For type I the treatment method of choice is conservative. For type III and IV a surgical treatment is necessary. Only the ideal treatment for type II remains unclear. There is no consensus in whether conservative treatment or surgical treatment will have the best outcomes. Surgical treatment mostly means a minimal invasive osteosynthesis. In our hospital we use percutaneous sacroiliac screw fixation with augmentation. The screws used are perforated and fenestrated, allowing for a correct placement an cement augmentation. Preferably, this procedure is performed in our hybrid operation theatre. An intraoperative CT-scan can be performed. This treatment method has proven to be a safe method in literature. On the other hand, surgical treatment, can have complications and risks, especially in the elderly. Conservative treatment consist of mobilisation and physiotherapy. But conservative treatment can be limited because of uncontrolled pain, resulting in a long immobilisation period. Immobilisation in the elderly leads to several complications. This research project will focus on early operative intervention to reduce the immobilisation period and its negative consequences.

Study design. Recruitment of patients will find place in our emergency department and our outpatient clinic. A lot of patients with a FFP type IIb and IIc present themselves at our emergency department due to immobilising pain. All the elderly patient who complain about sacroiliac pain will receive an CT-scan. Literature showed that an conventional x-ray is not sufficient to detect sacroiliac fractures. On the other hand a lot of patient are referred to our outpatient clinic by other hospitals in our region or even general practioners. All these patients, who meet our inclusion criteria, will be asked for informed consent. After informed consent has been collected, patients will be randomised to on of the groups. The first group will receive surgical treatment and the second group will receive conservative treatment. Randomisation is performed with a vending machine. Depending on which soda can comes out, patients are assigned to the groups.

Study intervention. As stated before the are two groups in this research project. Both treatment are standard treatments. Both groups will receive a standard set of co-interventions, such as adequate analgesics and intensive physiotherapy. At our hospital a geriatric trauma centre is established. After discharge or 4 weeks after trauma a first follow-up is planned. This is primary for the treatment evaluation. Then after 4 months are second follow-up is planned with a traumatologist, physiotherapist and a geriatric specialist. At this follow-up several benchmark test are performed, such as the DEMMI, EQ 5D and an accelerometer is explained and given out. Especially this accelerometer will tell us more about the amount of mobilisation or immobilisation at home. The last follow-up will be 1 year after trauma. At all follow-ups radiological controls are performed. These will be conventional x-rays. Only in specific cases, for example prolonged pain, an CT-scan will be performed.

Data and data management. The sample size was based on the expected difference between treatment groups in improvement on the DEMMI score between baseline and 4 months follow-up. Previous studies show that the minimal clinical important difference of the DEMMI score is 10 points. This results in a sample size of 68 patients, accounted for 10% loss to follow-up.

The statistical planned analyses are primary a pearson chi-squared or fishers exact test for categorical variables or a students t or mann-whitney test. However the primary outcome will be analysed using mixed linear models with random effects. The models will be compared usin Akaike information criterium. Missing data will be imputed using multiple imputation. All analyses will be performed using SPSS version 19 or higher. A p value <0.05 is regarded as being statistically significant.

Data is stored on the protected server systems of Hospital of Lucerne. Files containing anonymous data can also be stored on the personal computers of the investigators. Data is recorded on paper and digitally. Questionnaires are on paper, but all the measurements, as the DEMMI score of range of motion will be digitally. Participants can not be identified in the CRF (Case report form). Appropriate corresponding codes are only known and accessible for the investigators.

Upon simple request by the patient, he or she will immediately be withdrawn from the study and no further date will be recorded in the study. It is guaranteed that further treatment will be equal to standard care.