Malpractice claims after antireflux surgery and paraesophageal hernia repair: a population-based analysis

National AS and PEHR patient cohort data were obtained from the care register of the Finnish Institute for Health and Welfare between Jan 1st, 2010 and Dec 31st, 2018 [23]. All hospitals in Finland are mandated to annually report the details of every patient and the care given to the care register, with a reporting accuracy of 75–99% for common diagnoses [24]. Operations and procedures were reported using the Nordic Medico-Statistical Committee’s (NOMESCO) Classification of Surgical Procedures. For this cohort, cases involving AS and PEHR were identified using NOMESCO codes for AS and PEHR procedures (JBC00, JBC01, JBB00, JBB01, JBB90, JBB92, JBB91, and JBB93). From the Finnish Care Register data, we were also able to formulate a Charlson Comorbidity Index (CCI) for each patient [25]. For these patients, mortality up to 90 days was calculated from Finland’s national statistics on causes of death.

Additionally, data from all adult malpractice claims for AS and PEHR operations between Jan 1st, 2010 and Dec 31st, 2020 were collected from the national Patient Injury Centre (PIC) registry using the previously mentioned NOMESCO codes. Parallel to the judicial system, Finland uses a no-fault patient insurance system, which results in low-threshold reporting and the claiming of compensation without any legal advice. The PIC handles all the claims filed from both public and private health care systems, decides if an injury is compensable, and pays compensation accordingly. Patient insurance in Finland and the duties of the PIC are laid down in the Patient Insurance Act (948/2019), which also mandates that all care providers must have patient insurance. The PIC holds a national database of all filed patient injuries, including applicable patient records and the outcome of the claim. The most common criterion for a claim to be deemed compensable is its preventability: The PIC experts evaluate whether an experienced medical professional could have avoided the injury by choosing a different course of action [26, 27].

When analyzing PIC data, cases were separated into two categories. Those without a hiatal hernia and those with a type I hiatal hernia, which were classified as AS. Cases with type II, III, or IV hiatal hernias were classified as PEHR. Both AS and PEHR were included in the analysis due to similarities in their operations. However, the distinction was made because the decision-making process for whether to operate is very different for AS and PEHR.

A CCI was calculated for each patient. Records of all postoperative complications involved were collected, the complications were classified by the Clavien–Dindo classification for complications, and a Comprehensive Complication Index was calculated for each patient [28, 29].

This study is reported in line with the Strengthening the Reporting of Cohort Studies in Surgery (STROCCS) statement criteria [30]. As a retrospective registry study, no approval from an ethics committee was needed.

A total of 5734 AS and PEHR operations were performed during the study period. The mean age of the patient population was 54.9 ± 14.7 years and 59.3% (n = 3402) of the patients were women. The CCI was 0 for 58.6% (n = 3160), 1–2 for 34.6% (n = 1866) and ≥ 3 for 6.8% (n = 367) of patients. Overall, 432 (7.5%) operations were open and 5302 (92.5%) were laparoscopic. The intraoperative conversion rate was 0.56% (n = 30).

To estimate the number of reoperations, we identified 341 (5.9%) patients who underwent multiple AS or PEHR operations within the study period. Twenty-five (0.4%) patients had three or more operations. The index operation was used to evaluate patient comorbidities and was used in the comparative studies for statistical accuracy. Of these, 79.9% (n = 4384) were coded as AS, 20.1% (n = 1101) as PEHR, and 1.7% (n = 92) were coded as both. Annual rates of AS and PEHR are shown in Fig. 1.

The 30-day and 90-day all-cause mortality rates were 0.45% (n = 24) and 0.61% (n = 33), respectively. The 30-day mortality was 0.23% (n = 19) for AS and 1.36% (n = 15) for PEHR. The 90-day mortality was 0.27% (n = 12) for AS and 2.1% (n = 23) for PEHR. Of the patients who underwent PEHR operations, 12.8% (n = 200) had an ICD-10 code related to incarceration (K44.0) or necrosis (K44.1) of a hiatal hernia.

A total of 36 centers were identified, 11 (30.6%) of which had an annual volume of over 20 AS and PEHR operations, 11 (30.6%) performed 8 to 20 AS and PEHR operations annually, and 14 (38.9%) performed fewer than eight AS and PEHR operations annually. Of the operations, 68.8% (n = 3708) were performed in centers with a volume of over 20 AS and PEHR operations, 24.5% (n = 1320) in centers that had a volume of 8 to 20 AS and PEHR operations, and 6.8% (n = 365) in centers with fewer than eight AS and PEHR operations annually. Regional central hospitals performed 2656 (46.3%) of the operations, whereas university hospitals performed 2394 (41.8%) and other hospitals performed 684 (11.9%).

Logistic regression analyses for preoperative factors are presented in Table 1. Older age, CCI over 3, and open surgery are associated with higher mortality.

Between 2010 and 2018, 5734 AS and PEHR procedures were performed in Finland, resulting in 53 PIC claims, with an incidence of 1/1000 operations/year. Additionally, 7 claims were filed from 2019 to 2020 and analyzed. PIC claims were mostly filed after severe complications and comprised 50% AS and PEHR. By comparing the baseline data, PEHR and redo surgeries were overrepresented, and PEHR and AS showed different complication profiles, with patients filing claims after PEHR having a higher Comprehensive Complication index.

Comparing previous national reports of AS and PEHR, the combined rate of AS and PEHR decreased from 1400 operations in 2001 to 611 operations in 2018 [20]. In contrast to these reported numbers, the rate of PEHR increased during our study period from 60 operations in 2010 to 223 operations in 2018. Additionally, compared to the earlier national report on PEHR from 2001, the mean rate of AS and PEHR operations was 37.8 operations annually, whereas in our study, the mean rate was 137.6 operations per year [31]. These trends are in line with other studies of different populations [5, 6, 32]. The increase in PEHR might be explained by the increased usage of laparoscopy in more complex cases, as laparoscopic PEHR has less associated morbidity than open surgery and has been shown to be safe for the elderly and patients with comorbidities [33‐39]. The increase in popularity of PEHR may also be related to the rapidly increasing proportion of aging adults in the population population and increase in computed tomography imaging [40, 41].

In our cohort, the all-cause mortality rate after AS was low and comparable to that in similar studies [5]. The 30-day and 90-day mortality after PEHR were higher—1.36% and 2.1% respectively. This may be due to cases requiring urgent and emergency PEHR, which are known to bear a much higher mortality [42, 43]. Although other studies have stated that laparoscopic PEHR is safe in elderly patients, our multivariate regression model and univariate analysis of claim data showed that older age was associated with higher mortality and morbidity rates [44, 45].

Compared to the baseline cohort, the patients in the PIC group were slightly older and underwent more PEHRs. Of the operations described in the claims concerning PEHR, 70% had been performed in high-volume centers, which is thought to be associated with better outcomes [46]. Emergency PEHR operations were underrepresented in the PIC group, as only 6.7% were for emergency PEHR, whereas in the population data, emergency PEHR was involved in 12.8% of cases.

In the PIC patient group, reoperations were overrepresented, as 18.3% of the claims dealt with reoperations, whereas 5.9% of all patients had one or more reoperation after the index operation. This is in line with the notion that although considered to be safe and feasible, redo AS or PEHR has higher morbidity and mortality rates than primary operations [47, 48]. This highlights the importance of patient selection and informing patients of increased risks of adverse events, especially when considering reoperation.

Most PIC claims dealt with objective severe complications in both the AS and PEHR patient groups (Table 4). The patients who filed claims concerning PEHR had a significantly higher mean Comprehensive Complication Index than those who filed claims concerning AS. This can be attributed to increased frailty due to aging, as the PEHR group was older [49]. Additionally, the hiatus was reinforced with a mesh in 20% of patients in the PEHR group, which a recent meta-analysis by Angeramo et al. found to be associated with a higher overall morbidity rate than repair with only sutures [50]. Although the differences were not statistically significant, the profile of complications was different in the AS and PEHR groups. Perforations occurred more during PEHR than AS. This is in line with a previous institutional study by Zhang et al. consisting of 1223 foregut surgeries [15]. Dysphagia was more common in the AS group, although the difference from the PEHR group was not statistically significant. This might be explained by the higher rate of partial fundoplications in the PEHR group, as they are associated with less dysphagia (Table 3) [12, 51].

In the AS patients within the PIC group, preoperative diagnostic testing was not performed according to current guidelines, as only 43.3% of the claimants had pH surveillance done preoperatively (Table 2); although this has been recommended since the 1990s. [1, 4, 52] Also, preoperative manometry was performed only in 50.0% of AS patients in the PIC patient group (Table 2). These findings do not explain the observed adverse events, but they put the justification of the operation into doubt. Adherence to guidelines for the baseline cohort is unknown. In the randomized trial by Spechler et al., 79% of the enrolled patients did not meet the criteria for surgery, thus calling the extent of workup needed to find good candidates for antireflux surgery into question [2].

In previous literature examining the relationship between surgeon or hospital volume and patient outcomes generally favors higher volume [18, 53, 54]. In our data regarding all operations, 30-day mortality was lowest in medium-volume hospitals (8–20 per year), which can be attributed to selection bias, as more complex cases are referred to academic and other high-volume hospitals. Most of the PIC claims involved operations performed in high-volume hospitals, which is most likely due to the same bias.

Our study has several limitations. First, we were not able to obtain baseline data for all AS and PEHR operations performed from 2019 to 2020 due to changes in laws concerning the secondary use of health and social data. Second, as for all retrospective analyses, ours is prone to selection bias and, therefore, cannot conclude causality. Third, the Care Registry does not provide information on potentially significant confounding factors like obesity, smoking status, and prior abdominal surgery. Additionally, we were only able to assess the volume of the hospitals, not the volume of single surgeons. Additionally, in Care Registry data, AS and PEHR procedures are differentiated based on different NOMESCO codes instead of the actual sizes of the paraesophageal hernias so some cases may be miscoded. We were only able to report overall mortality, not disease-specific mortality. Although the threshold to file a claim is low within the national PIC system, it is patient-driven, and therefore not all complications are reported. PIC data may also be affected by sampling bias, which might reduce the generalizability of the results [55]. From the patient injury data, we could not assess the severity of symptoms, usage of PPIs, or quality of life. Because of the timeline of the claim process, some claims concerning operations performed during our study time may not have been filed before these data were collected.

This study highlights a significant change in the trends of foregut surgery, with the focus shifting from AS to PEHR at the national level. This transition is also evident in the number of malpractice claims, with relatively more claims associated with PEHR due to the complexity of the surgery, older patient age, and medical comorbidities. Nevertheless, both AS and PEHR are generally safe during the era of minimally invasive surgery, and patient injury claims are rare. As the aims of both procedures are to alleviate symptoms and improve quality of life, the occurrence of severe complications should be extremely low.