Subscribe via RSS Feed

Hammering away on the peritonsillar abscess theme…

April 22, 2019

Lloyd Gordon sent me this article. It is not a study regarding POCUS guidance for peritonsillar abscess (PTA). PTA drainage procedures in this study appear to have been done by ENT surgeons using old school blind technique. But it is a negative study. So good to tackle nonetheless. The article is by Battaglia et al and was published last year in the journal, Otolaryngology–Head and Neck Surgery. Here is the PubMed link. Here is the summary.


Objective: Compare initial medical therapy (MT) vs initial surgical therapy (ST) with MT for PTAs

Null hypothesis: Complication and failure rate equivalent for either form of treatment

Mention is made of past studies showing no difference in the failure rate when comparing different surgical procedures such as I&D and needle aspiration (range: 6-20%).



Kaiser HMO hospitals in Southern California, specifically 12 centers that were known to have adopted a protocol of providing MT initially without ST for PTAs AND 7 centres that were known to have continued with the traditional protocol of providing ST initially with MT as well.

January 2008 to January 2013

According to treatment location analysis, 92% of all abscesses treated at the 7 service centers doing primary ST between 2008 and 2013 had a procedure, while 92% of all abscesses treated at the 12 other service centers doing primarily MT were exclusively treated with MT and not ST…ST was reserved for treatment failures.

Inclusion criteria

  • Adult or pediatric
  • Clinical diagnosis of PTA by EP
  • Had MT exactly as per protocol at one of the centres mentioned above

Power calculation

Determined that at least 84 patients would be required in each group for the various outcome measures

Outcome measures

The following were recorded from charts on initial presentation, 1-2 hours after treatment, and at 1, 2, 3, 7, and 42 days after presentation:

Pain scores (one point difference being significant), fever/temperature, trismus, uvular deviation, toleration of liquids/solids, number of days until PO intake, volume of pus during procedure, dyspnea, bleeding, number/volume of opioid prescribed, average number of sore days, number of clinic visits, days off of work, failure rates, and complication rates.

Complications included aspiration, bacteremia, hemorrhage, mediastinitis, parapharyngeal abscess, retropharyngeal abscess, sepsis, syncope, palpitations, and weakness.

No mention if the above data was collected at return visits at these set days, via phone interview, or only if the patient happened to come to one of these Kaiser facilities

Medical treatment protocol (developed from a prior pilot study in 2008)

  • At the time of ED diagnosis:
    D5 ½ NS,  1 L bolus
    Dexamethasone 10 mg IV
    Ceftriaxone 2 g IV
    Clindamycin 600 mg IV
  • After discharge: Clindamycin 300 mg QID for 10 days
  • Patients are observed in the ED for 1 to 2 hours after medication administration.
  • If patients are not significantly improved with the ability to take liquids by mouth, they undergo needle aspiration or incision & drainage
    Patients are followed up by ENT the next day. If patients are significantly better, they are discharged. If they are not significantly better, they are either surgically drained or given another round of IV fluids, dexamethasone, and ceftriaxone. After the 2nd round of medications, if patients are not significantly improved with the ability to take liquids by mouth, they undergo surgical drainage.


Southern California Kaiser database searched.

Found 6782 patients diagnosed clinically with a PTA by emergency physicians, Jan 2008 – Jan 2013

Of these, 6132 (90%) treated without ST; 650 (10%) were treated in ED with ST and MT

Narrowed down to the 1747 patients with a PTA, who according to the coding/documentation, had “uncomplicated” abscesses and received therapy exactly according to the MT protocol.

Then reviewed those 1747 patients charts to see who received the same MT regimen but who also received standard surgical drainage at one of the 7 service centers where ST was routinely performed; 96 such cases were identified

Outcomes for these 96 ST patients were then compared to 211 randomly selected out of the 1747 MT patients treated without drainage at the other 12 service centers where MT was the sole primary treatment.

  • No difference in baseline characteristics such as age, sex, level of pain, fever, ability to take PO.
  • ST group significantly more likely to refill narcotic prescription
  • ST group had significantly more sore days and more days off work

Treatment failures

  • 17 patients (8.1%) in the MT group had a procedure within 42 days
  • 6 patients (6.2%) in the ST group had a 2nd procedure within 42 days

Patients with trismus at presentation (as surrogate to see if ST group patients were sicker than MT group patients at baseline)

  • 39 (18.4%) cases of trismus in the MT group; 44 (45.8%) in the ST group
  • BUT only 3 patients (7.7%) in the MT group ended up needing a procedure due to treatment failure; and 3 of the patients in the ST group needed a 2nd procedure


On the face of it, these results suggest that PTAs are better treated initially with medical treatment only and that drainage procedures should be reserved for cases that do not improve with medical treatment.

One significant limitation with this study is that we don’t know really know the diagnosis. The diagnosis in this study was a clinical one. If memory serves, in the Soft Tissue presentation at EDE 2, Ben Ho quotes a 75% rate of accuracy for PTA diagnosis by ENT surgeons! Emerg docs are no better. Bottom line is that Emerg and ENT docs should not try to drain a PTA before knowing that there actually is one. Why put the patient through that otherwise.

That begs the question…Did some of these patients not have a PTA at all? Maybe some only had peritonsillar cellulitis? Maybe that’s all that most of them had? Do patients with peritonsillar cellulitis do worse if you stick a scalpel in their tonsil? You bet!

Then there is the issue of only including “uncomplicated” PTAs. What does that mean? Could it mean that none of these patients were admitted? Because it looks like none of them were. Most of my PTA patients in Sudbury are at least borderline for admission…so maybe a different population. So, in the patients who actually did have a PTA in this study, maybe some of them only had a really small one? In that case, I agree that MT alone may be all that is needed.

We also don’t know who had an I&D vs who had a needle drainage. It would need to be studied as well, but it stands to reason that the former would lead to more discomfort and a bit slower healing time. We also don’t know if all the ST cases were successful and had at least some pus drained? As an aside, most PTAs are quite liquid and amenable to needle drainage.

By the way, I should mention that the authors acknowledged all or most of the limitations mentioned above.


If you don’t know if your patient has periorbital cellulitis or a mild PTA and you don’t have access to imaging (POCUS, elective ultrasound or CT), give IV antibiotics, IV fluids, and IV dex, as indicated.

Better yet, use POCUS! See related/recent blog posts below.

Future studies

Really this study enrolled patients who might have had a PTA which, if present, was uncomplicated (i.e. not that bad, possibly small in volume). Future studies should confirm the diagnosis by some form of imaging, as well as confirming the PTA size, before we can better know the role of drainage, initial vs 2nd line.


Filed in: Articles • Tags: ,

Comments are closed.