Approximately 10% of our almost twelve thousand respondents stated that they personally ever had experienced an AE, and 15–19% that an AE had affected their near relatives. Taken into consideration that this is a life time cumulative incidence, it is a considerably lower incidence than would be expected from data in small samples population studies, partly recorded during limited time periods [10–13]. The lower incidence of AEs found in our study can be explained by recall bias and the fact that most AEs are minor and cause only temporary symptoms [2, 4–6, 9]. Major adverse events may be remembered for life, but minor events may be remembered only if occurred recently. On the other hand, our study population obviously includes some first degree relatives. Therefore, some respondents may report the same adverse event occurring in the same relative. This would, to a minor degree, inflate the estimates of prevalence of adverse events reported for relatives.
The strength of our study is the large representative sample obtained from a general population, and a high response rate. Further, this study, in contrast to many other studies of medical adverse events or medical errors, also includes the public’s experience from contacts with the primary health care, as more than four out of five of our respondents reported to have consulted a GP last year and consequently the majority of the reported AEs were related to the GP.
The present study, as well as other studies exploring incidence and causes of adverse events, has an important methodological problem: The identification of a medical error, a mishap or an adverse event is a subjective process. This is true for lay people as well as health professionals. Attempts to reduce the problem have been done by introducing definitions [1, 15], examples [9, 12] and improved taxonomies , but the problem still persists. Some authors have focused on challenges especially related to primary care [17–19]. In our study, we have not defined the concept “adverse event”. We have only asked whether the respondents have experienced that insufficient examination or treatment has had “serious consequences”. Their answers therefore include all kinds of unexpected outcomes. A deeper exploration into these challenges needs further studies.
We found that respondents who had felt persuaded to accept an examination or a treatment they did not want, and respondents who had found it difficult to be referred from a GP to specialist health care, had a markedly increased likelihood to report AEs, affecting themselves as well as their relatives. This probably reflects aspects of the patient-doctor relation: A more consumer-like attitude from patients, with increased expectations of easy access and a perfect outcome, combined with strengthened patient rights, and increasing demands for service will inevitably increase the occurrence of perceived AEs.
As an indicator of communication problems between patient and doctor, we asked the respondents if they understood what the doctor said during the last consultation. We found that such communication difficulties strongly increased the probability for reporting AEs. This is in accordance with other studies [20, 21], and indicates that doctors should aim for better communication skills .
Respondents reporting to have experienced an AE, occurring to themselves or their relatives, stated that their GP was responsible for the AE in about 60% of the cases, hospital doctors in about 40%, and the system (failing routines or communication) in 20%. This distribution of perceived responsibility probably in part reflects the frequent contact between the respondents and the primary health care, but also indicate that patients interact and relate more to individual health professionals, in particular doctors, than to systems. Our results are in line with studies by Blendon et al.  and Northcott et al. , who found that members of the public, as well as doctors, were more likely to place the responsibility for medical errors on individual health professionals than on the institution involved. Doctors, management and superior authority who are confronted by patients or public to discuss AEs should consider the distribution of perceived responsibility found in our study, even if the distribution may be felt unfair.
When strategies for quality improvement in health care systems are elaborated, the prevailing view is often that adverse events primarily are caused by failures of institutional systems. This perception of reality is not supported by the present population. Positive effects of system based interventions to improve patient safety have only modest supporting evidence . In their daily work doctors may tend not to adhere to protocols, and to be reluctant to use standardised routines . In patient safety work, where avoiding “blaming and shaming” and search for scapegoats is part of the recommended strategy, the personal responsibility of doctors and of other health professionals still should be emphasised.
Although about 40% of the respondents having experienced AEs personally considered complaining about the events, only about 10% in fact did so in writing. It has also been shown by others that AEs rarely lead to complaining, and very few medical errors result in medical-malpractice claims . Norway and other countries have several complex and often uncoordinated systems for recording and handling of patient complaints. In Norway, a country with about 5 million inhabitants, a so called no-blame system, “The Norwegian System of Compensation to Patients” , was established in 1988. This system, intended to be a low threshold system, receives compensation claims from patients treated in both public and private health services. In 2010, the system received 4352 claims from patients who had experienced an AE. Seventy eight % of the patients were older than 30 years of age, and 32% of them achieved economic compensation.
Further, it should be noticed that only approximately 50% of the respondents who reported to have complained in writing in the present study, reported to have experienced AEs personally, and about 30% of them that AEs had occurred to their relatives. This means that many complaints are caused by other factors than experienced AEs, as communication problems or substandard professional behaviour. Altogether, this implies that patient complaints are of limited value as a basis for efforts to reduce the incidence of AEs.