This retrospective study describes adverse events after orthopaedic surgery. The following findings were noted: the high number of adverse events caused by infections after orthopaedic surgery and the high degree of serious disability after adverse events in connection with spine surgery. Another important finding was the relatively high number of adverse events causing permanent disability in the two large groups involving primary knee and hip prosthetic replacement. The finding that hip and knee replacement and intervertebral disc surgery were the most common procedures resulting in sustained claims is in accordance with what has been reported in a small American study .
The nature of the national systems for compensation of adverse events may influence the extent to which review of claims data can be used to learn more about adverse events. Claims data in the United States are primarily used by litigation managers, attorneys, and others for determining legal liability. Previous studies have shown little relationship between errors and malpractice claims . The no-fault system in Sweden, in contrast, does not place the responsibility for a medical error on an individual practitioner and may reduce barriers to filing for compensation and increase the probability that an error is disclosed . The relative rate of claims in Sweden exceeds that of countries with a litigation system, suggesting that its reporting system may contribute to making claims data more useful for identifying and learning more about adverse events.
In Sweden, a third of all patient claims involve orthopaedic treatment, whereas in the United States the percentage of orthopaedic claims is around 14% of the total claims . Our finding that orthopaedic surgery had the highest share of claims among all specialities at the insurance company is in accordance with what has been reported before in a study from another Nordic country with a no-fault system .
Jena et al.  reported malpractice claims according to speciality in the United States and found that neurosurgery, thoracic-cardiovascular surgery, general surgery and orthopaedic surgery were the specialities with the highest probability of claims. In Sweden, the following specialities have the highest probability of a claim: hand surgery, orthopaedics, cardiothoracic surgery and neurosurgery .
The high rate claims in connection with orthopaedic surgery might be due to high patient expectations or because complications are more obvious than in other specialities. Knee and hip prosthesis surgery are considered by the public to be very safe routine procedures and if the result is not what the patient expected, it may result in a claim. Wong et al.  reported that the reasons for adverse events in orthopaedics are multifaceted; for example, communications failure, equipment and/or instrumentation problems in the operating room, improper technique and/or physician impairment. However, these reasons are no different for other surgical specialties.
The number of patient claims and adverse events in relation to hospital admissions varied tenfold between different age groups in this study. In the most elderly group (> 80 years), fewer adverse events were found despite the fact that this age group is the largest in most orthopaedic departments. This could be explained in two ways: either elderly patients undergo less complex procedures with lower risk or they are less prone to claim economic compensation. The high rate of claims in the group aged 20-59 years can probably be explained by a higher tendency to report injury, most likely related to loss of income during working years.
This study reveals that almost 1% of all orthopaedic patients receive compensation for adverse events assessed as preventable. Even though the number of patient claims is increasing slightly year by year, the actual number of adverse events is probably much higher, as indicated by the results of a national study showing that 8.6% of patients experience preventable adverse events .
The frequency of hospital-acquired infections and sepsis after orthopaedic surgery in this study is in agreement with findings in an Australian study on adverse event in general surgery and orthopaedic surgery . A high rate of hospital-acquired infections and sepsis is also reported from the United Kingdom in studies on adverse events [3, 4] but there is a paucity of studies on the frequency of infections as a cause of adverse events in orthopaedics. In our study, 22% of the patients were awarded compensation on the basis of infection, which is higher than previously found in a Norwegian study (8.1%) . At least in Sweden, retrospective analysis of patient claims data seems to be able to identify and provide additional information on the consequences of hospital-acquired infections. When we investigated 113 cases of injury leading to serious disability and compared these injuries with a mandatory hospital-based sentinel event reporting database, we found that none of the 19 claims involving health care-associated infections that led to permanent disability were reported by the providers . Furthermore, when reviewing patient claims after spine surgery, health care-associated infections were the second most common adverse events leading to compensation .
Procedures to decompress spinal nerve roots and the spinal cord result in adverse events more often than other surgical procedures . In our study, nearly one-fourth of the patients suffered serious disability after spine surgery. In a study with the aim to describe and analyse the outcome after spine surgery, claims data were compared from the County Councils' Mutual Insurance Company with data from a national register and medical records. It was found that dural lesions were a common, but underreported, complication and an important reason for problems contributing to high levels of disability .
Our study is a retrospective analysis of the contents of the County Councils' Mutual Insurance Company database. The variables used by the insurance company are primarily designed for the reimbursement procedure and not for medical analysis. The data set is complete only for patients who received economic compensation. Despite these shortcomings, we consider the database to be an important source of information on adverse events. The study is a survey of a whole range of orthopaedic injuries but has nevertheless identified some risk groups. In order to identify underlying causes of adverse events in connection with hip and knee replacement, it would be interesting to compare claims data with data from national registers and medical records.
We conclude that patients undergoing spinal surgery were subjected to high risk of being severely injured and that these patients also experienced a high degree of serious disability. The most common adverse event was related to hospital acquired infections. Claims data obtained in a no-fault system have a high potential for identifying adverse events and learning from them.