Page 106 - International guidelines for groin hernia management

Hernia

Table 5 Pros and cons with registry studies and randomized studies
Pros Cons

Registry-based High external validity Requires a limited contribution of a large number of
studies Includes all patients at aligned units surgeons
Involves many surgeons with varying level of skill and experience Recurrence rate are replaced by re-operation for
recurrence
Reflects routine clinical practice
Generally lower rate of follow-up than in RCT
Provides separate data from participating hospitals and aggregated
Low internal validity if not data are monitored
data for all participating units
Requires a limited contribution from all surgeons
Excellent tools for observing changes over time
Includes documentation and adjustment of several confounders
Permits post hoc subgrouping of patients at high risk
May investigate even rare events
Power increases over time
RCT High internal validity Specified inclusion and exclusion criteria limit the
Allows for comparison of methods of repair under standardized external validity
study conditions Inclusion of all consecutive patients is difficult
Simple statistical analyses with comparative methods Results are mostly obtained by a limited number of
Can prove the impact of a specific change in treatment on a experts under optimal conditions
specific outcome in a specific setting Extensive contribution from participating study
Short-term rates of recurrence and chronic pain can be determined investigators is required
Follow-up for more than a few years is rarely possible
Focus on a single primary endpoint
All confounders are usual not considered
Post hoc subgrouping usually is not possible or justified
Usually insufficient power to detect rare events

For the reasons cited above, RCTs and registries should studies can shed light on rarer events and conditions like: IHs
be considered alongside one another when evaluating in females, femoral hernias, serious complications and
various aspects of hernia repair. 153 mortality. 155, 156 It has been demonstrated the results
Ideally, a registry should follow patients from initial abstracted from Danish and Swedish databases have changed
157–159
inclusion event to death. Provided consent is obtained to clinical practice nationwide.
use personal identification numbers, patients can be tracked Use of the checklists from the CONSORT statements,
within a healthcare system and all subsequent encounters the STROBE curriculum and the RECORD statement are
(e.g., reoperation) recorded. 145, 146, 154 It is also possible to highly recommended to improve reporting quality for
link registries of various types to detect and analyze risk RCTs and observational studies. 144, 160, 161
factors contributing to unfavorable outcomes. The 2012 EHS consensus meeting also spawned rec-
Coverage and data validity are crucial for registry studies. ommendations for reporting outcome results in abdominal
If a registry enrolls nearly all hernia patient encounters, the wall repair across different study types. 162
risk of skewed patient selection is minimized. Additionally,
care must be taken when entering registry data since incor-
rect or missing data limit a registry’s soundness. These fac-
tors influence the external validity or generalizability of
conclusions reached in studies involving registry data and
patients. In a perfect world, registry data and the conclusions
about those data would exactly match the world outside the
registry. This ideal is unlikely to be realized, but regional and
national registries do include enormous data sets. Studies can
be performed about time trends for repair methods, materials
Figure 8. The potential coverage of patients operated on
used, anesthetic type, patient gender and others topics. In
for an IH reported in a national or regional register com-
contrast to studies performed at a single institution, registry
pared to a randomized control trial.

123