No patient is the ever same and this adage goes for shoulder dislocations. If you use the same technique every time there is going to be situations where you are simply using the wrong technique. That’s why I take an algorithmic approach to the patient in front of me. This includes addressing whether this patient needs an X-Ray, what type of injury they might have, whether this patient need sedating, and pre-planning what to do if Plan A fails. There are anatomical reasons why no technique works every time on every patient and it is important to be skilled in several methods so you can use the best method for each presentation. The techniques discussed rely upon the relative movements of the humeral head and the glenoid fossa of the scapula. Used correctly, these techniques require little or no force.
Here’s the generic algorithm, once you have mastered a few techniques you can adapt this to your particular skill set. For example, if your favourite zero position technique is FARES, then that fits into your own personal algorithm – my algorithm appears below the generic one.
I'd also recommend declaring your plan to your patient and your team. 'Preparing to fail' is a great way to make good decisions in evolving situations especially when things aren't going to plan. Saying to your patient "I'm going to try a gentle technique using positioning and massage. Your shoulder may pop back in just with that, if not we are going to stop, get an X-ray, and get a bit more information before we try the next technique." This allows you to maintain the confidence of your patient (and yourself) if Plan A doesn't work, and stop you from spending too long on a technique that doesn't suit the patient or the injury.
For more detail on whether to X-Ray check out the FAQ section
And if you happen to be working on a football or ski field, here’s my field algorithm.
The short answer is that whatever your dislocation technique skill set is at baseline, if you can use the available data to improve your decision making at each step, you will increase the probability of success.
The long answer is below!
For every shoulder dislocation scenario there will be a decision to make at each point. These decisions will be based upon information that may range from being fully available and requiring analysis, to unavailable or partially available. There will need to be a risk analysis for your decision once the availability of data has been assessed. Most of the decision points have not been settled by rigorous scientific experiments so expertise needs to be used alongside experience, so the process below will use estimates that will vary based upon your own skill level with each technique. Due to the various factors involved in each dislocation, the ‘what do I do next?’ questions are all complex and prone to cognitive bias with regards to risk assessment. Going with gut instinct or using the same technique you always use is not going to helpful when to encounter a different pattern of patient/injury factors that renders your go-to approach unsuitable. I’ve used a case example below to illustrate the various factors to consider for each scenario, using the algorithm to create an individualised plan for your patient. All prediction numbers are estimates based on expertise rather than hard science, and as you will also see from the footnotes, there are many unanswered questions that, once properly researched, could allow us to improve our predictions and therefore our individualised patient plans.
“I always use Kocher’s technique and have a great success rate of 90%”
Operator’s self-assessment of likelihood of success with Kocher’s technique using intuition – 90%
After assessment of all data available in this scenario:
Likelihood of success with Kocher’s technique – 30%
Likelihood of success with Zero position technique – 55%
Likelihood of success with seated scapular manipulation technique – 60%
Likelihood of success with prone scapular manipulation technique – 90%
Patient factors - 58 year old male, overweight but otherwise well, cooperative, pain controlled (3/10) after IV morphine
Injury factors - mechanism of fall onto shoulder, X-ray shows greater tuberosity avulsion fracture (GTa#) and sub-glenoid dislocation
Operator factors – Previously worked in Ski-fields and experience of over 50 successful reductions. First line is Kocher’s, uses morphine if needed and if unsuccessful moves to sedation and traction.
 Research question – This patient already has an XRay. How accurate are emergency personnel at predicting anterior (sub-coracoid) or antero-inferior (sub-glenoid) shoulder dislocations based on mechanism of force and clinical assessment?
From an anatomical perspective Kocher’s is not the best choice for this scenario, despite the high level of proficiency held by the operator. Kocher’s technique was designed for sub-coracoid, not sub-glenoid dislocations. This does not mean that it neverworks for sub-glenoid techniques. It can sometimes work with sub-glenoid techniques, if: some adduction is possible to appose the articular surface of the humeral head to the glenoid rim; the humeral head does not wedge beneath the glenoid rim, and; the first manoeuvre (external rotation) allows the articular surface of the head to roll over the rim. Let’s estimate this at 30% of sub-glenoid cases. The second/third manoeuvres are definitely not going to work in this case as there is a floating greater tuberosity - a fixed one is needed during these parts of the technique.
The operator’s technique proficiency in Kocher’s is high, but their experience in ski-fields is more likely to be based on younger patients with a higher proportion of sub-coracoid + Bankart’s lesion injury patterns. If we assume that their performance of Kocher’s technique is perfect (100%) then the chance of them being successful with it is estimated as 30%, with reduction occurring during the first manoeuvre (external rotation).
This operator may still decide to attempt Kocher’s first manoeuvre (external rotation) and if that works – great. It is a fast and easily tolerated technique, and avoids direct involvement of the injured greater tuberosity in this case. With the extra information they have, they can explain this approach to the patient – “I’m going to try a gentle manoeuvre that may work quickly. Based on your injury type it may not, so we will then move onto a different option. Relax and let me know if it is too painful.” The expectation of the operator is now that the probability of this first attempt (Kocher’s first manoeuvre) working is 30%, but worth trying as it is a low pain, safe option for the patient.
They definitely won’t progress to the second manoeuvre (flexion) of Kocher’s if the first is not successful as there is low likelihood of reduction (in the absence of an intact greater tuberosity), and an increased risk of hurting your patient by apposing the fracture site and the glenoid rim. They also won’t attempt external rotation if the patient can’t adduct as this suggests wedging of the humeral head and a very low likelihood of reduction from this position.
They prepare a plan B based on the knowledge that there is a free floating GTa# and decide between zero position and a scapular manipulation options. We’ll assume perfect technique performance again from the operator, and extremely low risk of shaft fracture (GT avulsion rather than impact.
There is a risk that the free floating GT# forms an obstruction to the returning head (estimate increase 10% failure), or that the GT# origin on the humerus rests against the glenoid rim at the proposed point of reduction (25% estimate increase in the failure rate EIFR) causing pain and spasm. Both of these issues result in a failed reduction attempt. The patient is cooperative but in some pain from the fracture and haemarthrosis, this results in some residual spasm that may not be able to voluntarily overcome to allow reduction (EIFR 10%). Probability of successful reduction using zero position technique in this case is 55%.
This technique keeps the fracture point clear of the rim during reduction (nil EIFR). Again, the patient is cooperative but in some pain from the fracture and haemarthrosis, this results in some residual spasm that may not be able to voluntarily overcome to allow reduction (EIFR 10%). Starting with the humerus in forward flexion keeps the GTa# away from the point of reduction for the humeral head over the glenoid rim. In the seated position this patient’s heavy arm increases capsular stretch and displacement of fracture fragments (proximal humerus sags) causing excessive pain and spasm. This, along with inferior displacement of the humeral head (suboptimal for reduction) increases risk of failure by 30%. Even performed perfectly, the limitations from the seated starting position and pain/spasm from haemarthrosis result in an estimated probability of success of 60%.
Starting with a prone position with the arm in weighted flexion removes the inferior displacement of the humeral head and excessive capsular stretch (nil EIFR), but residual pain from fracture fragments and haemarthrosis leave an EIFR of 10%. Performed perfectly, the estimated probability of success of prone scapular manipulation in this scenario is 90%.
 Research questions - How successful is Kocher’s technique in sub-glenoid dislocation with GT# (first manoeuvre only) or no GT# (full technique)? I don’t think it ethical or useful to attempt manoeuvres 2 and 3 in a case with a GT# as there is no real expectation of reduction success, and a risk of inflicting unnecessary pain or further injury.
 Research question – What is the understanding of correct Kocher’s technique among emergency personnel?Data points would include; starting position of humerus, use of traction, exclusion criteria.
 Research question - What % of GTi# have an underlying shaft # on CT scan. A practical approach being that all GTi# should have CT scan. Increased risk of underlying # with either moderate force mechanism (fall) in older patient with relative osteoporosis, or with higher forces required to generate # in younger or male patients (denser bone).
 Research questions – What is the percentage of GTa# and GTi# in sub-coracoid and sub-glenoid dislocations? What is the success rate of a zero position technique with/without GTa/i#?
 Research questions – To what degree does the presence of a fracture (GTi#, GTa#, bony bankart’s, distant) and haemarthrosis (ultrasound) affect pain scores in acute shoulder dislocations?
 Research question - Does the presence of a # or haemarthrosis determine ‘cooperation with reduction attempt’ in acute shoulder dislocations?
If your usual approach is ‘Kocher’s for 90%, and sedation and traction for the remaining 10%' then you are not using all the available information and risk unnecessary pain during reduction techniques, or an avoidable sedation.
Note that the information available for the estimated data points is not absolute as there has not been research into specific sub-type injuries and the operator proficiency and patient factors vary from case to case. The footnotes collect some questions that have yet to be answered which could provide more accurate estimates to these predictions.
Kocher’s can sometimes work in sub-glenoid injuries, this can sub-divided further depending upon the injury type and the manoeuvre used. Assuming that the patient is cooperative and comfortable (pain/spasm), they are able to fully adduct the humerus, and the humeral head is not wedged beneath the rim, then Kocher’s can be attempted. All 3 manoeuvres can be used if the humeral head is intact; only the first manoeuvre (external rotation) can be used if the humeral head is damaged (GTa# or GTi#); an old Hill-Sach’s lesion without significant impaction could theoretically provide the fixed point to move between manoeuvre 2 and 3 but there really are other techniques that are more likely to succeed in this scenario.
Based on all the available data, this operator’s individualised approach for this scenario now looks more like this:
Plan A - first manoeuvre Kocher’s (30% likelihood of success),
Plan B - prone scapular manipulation (90%),
Plan C - sedation plus scapular manipulation (patient fully seated or angle based on airway management during procedure), 100%.
Here’s a tip for knowing when your zero position isn’t working and you need to try a different approach. If you can feel the scapula move in rigid tandem with any movements of the humerus once you are in an abducted position, and your patient is relaxed and cooperative, then you have a case where the head is wedged under the glenoid rim. Attempted movements include further abduction, external rotation or anterior/posterior, and you will feel the scapula move as if it is stuck to the head. You are unlikely to be able to coax the head into sliding out and will need to either sedate your patient and see if removal of all muscle spasm is the issue, or try an alternative approach. Moving the humerus into a forward (anterior) elevation position while an assistant keeps the tip of the scapula pushed medially, allows the head to dislodge from the wedged position and either reduce immediately, or move into a more free sub-glenoid position where you can re-attempt a zero position. The other option is formally moving your patient into a prone position and performing scapula manipulation. The take home message is that if you are stuck – stop, reassess why you are stuck, use the additional information you have gained based on your first attempt to be flexible and alter your plan.
Sometimes you have to manage situations in suboptimal environments. I tried my first spartan race this weekend with the team from my local gym @fitnesskickflemington and one of my team mates Jason popped his shoulder out mid race and somehow kept going to the finish. I often talk about how it is not important where you stand as an operator or whether the patient is sitting, lying, prone, or hanging from their feet as long as you know where the scapula is compared to the humeral head. In this case, Jason had a medal round his neck and a victory beer in one hand, so I reduced his shoulder standing in the field, flanked by men in capes.