Population health and orthopaedic surgeries

By | January 15, 2020

Both knee arthroscopy for partial meniscectomy and shoulder arthroscopy for acromioplasty were very common procedures few years back in orthopaedic surgery. During the first two decades of this millennium extensive amount of research has been published showing that the effect of these procedures to treat knee and shoulder pain and function is quite weak. Despite of this, the debate is still very active around these topics. While many practicers and researchers believe that these procedures are without benefit for patients, proponents are also active questioning the validity of these sham-controlled RCTs.

So, let’s take a closer look to Finland. They (we) a national health insurance system meaning that government pays the healthcare for every citizen. This means that everything done at the hospitals is recorded and registered. Numerous studies have been published about different surgical procedures performed in Finland (eg this, this and this). Those of most interest are studies by Mattila et al. (2016) and Paloneva et al. (2015) both published in Acta Orthopedica. Mattila et al looked at annual incidences of knee arthroscopy for osteoarthrosis (OA) and degenerative meniscus from 1997 to 2012. Paloneva et al looked annual incidences of acromioplasty from 1998 to 2011.

Many large and national health surveys has been done in Finland to get some insight to the overall population health. Main data sources here are Terveys 2000, Terveys 2011 and Finterveys 2017 surveys. These surveys has included structured questions about physical and psychological symptoms and complaints. These surveys have included answers from 8000, 8000 and 11000 persons, respectively. Main interest here are the following questions:

  • Difficulty in walking due to knee problems during last 30 days (Yes/No)
  • Difficulty in arm abduction during last 30 days (Yes/No)

Let’s match prevalence of walking difficulty to annual incidences of knee arthroscopies between 2000 and 2017.

Incidence of knee arthroscopy and walking difficulty.

It’s quite clear that the prevalence of walking difficulty has increased steadily. Since 2008 the incidence of knee arthroscopy has decreased clearly. Shaded area and dotted line is raw extrapolation from 2012 to 2017. Has the decreased incidence of knee surgery caused an increase in walking difficulty? Very hard to say. On the other hand, the incidence of knee surgery remained very high during the early 2000s. In 2006 almost 0.4% (400 per 100 000) of Finnish adult population had knee arthroscopy due to OA or degenerative meniscus. That’s a lot! Incidence of knee arthroscopy has been three times higher in Finland compared Sweden. Despite of very high incidence of knee surgery, prevalence of difficulty in walking actually increased from 2000 to 2011. Assuming that knee arthroscopy for OA or degenerative meniscus had some kind of an effect, shouldn’t the prevalence of walking difficulty decrease or at least remain stable? Again, very hard to say. It depends very much on the proportion of degenerative conditions in all reasons resulting to walking difficulty.

How about shoulder surgeries? Let’s match the annual incidences of acromioplasty to the prevalence of difficulty in arm abduction.

Incidence of shoulder arthroscopy and arm abduction difficulty.

Temporal trends are again very interesting. During early 2000 the incidence of acromioplasty increased rapidly. At the same time the prevalence of arm abduction difficulty slightly decreased. Are these trends reasonable assuming that acromioplasty has some kind of treatment effect in shoulder impingement? Furthermore, the prevalence of abduction difficulty has remained constant even though incidence of surgery have clearly decreased. Therefore the temporal trends in acromioplasty and prevalence of abduction difficulty don’t seem intuitive or logical.

There is lot of uncertainties in this analysis. Like, what is the proportion of OA, meniscal tears and shoulder impingement in all causes of these complaints? We also don’t no whether volume of surgeries is a marker for condition or vice versa and how causal paths would work if there was indeed a correlation between population health and these surgeries. Based on this data many different views can be taken. Still, I’m inclined to believe that if knee arthroscopy or acromioplasty had any kind of an effect in symptom relief, the trends observed here would be different. But, if proven otherwise I’m happy to change my perspective!

Leave a Reply

Your email address will not be published. Required fields are marked *