Seven papers, covering a total of 209 operations to remove the coccyx, were analysed. Seven of the coccygectomies were carried out to treat coccyx pain following earlier surgery to fuse parts of the spine. None of these were successful in relieving pain. Of the remaining 202 operations, 168 (83%) were successful. In a group of patients who had coccyx pain of unknown origin, seven out of twelve (58%) had a successful outcome. This is a lower success rate than in other groups, though the numbers in this group are too small to draw definite conclusions. The results also suggest that 1 or 2 % of patients may have increased pain after surgery.
This summary is based on seven trials of coccygectomy for the relief of coccyx pain published in the past twenty years [1-7]. I obtained full copies of these papers. I excluded two further studies, one written in Russian and one in Polish, which were available only in abstract [8-9]. The first of these abstracts did not state the success rate and the second contained a contradiction (presumably a mistranslation or typographical error). The seven full papers covered a total of 209 operations.
The studies were carried out in different ways and had different criteria for selection of patients and for deciding if an operation had been successful. Papers and  were planned trials using consistent criteria for the selection of patients. The other papers [3-7] reported retrospective studies on coccygectomies carried out over periods from 6 to 24 years. Paper  included operations carried out as long ago as 1955. Despite the differences between the studies, they produced broadly similar results, and some general conclusions can be drawn.
All of the studies excluded patients who were found to have cancer or in whom the source of the pain did not appear to be the tissues around the coccyx. In all cases (except one of the cases in ) 'conservative' treatment was tried before surgery. Papers  and  reported that in their cases 'conservative' treatment always included injections of corticosteroids, but this was not necessarily the case in the other studies. Two studies applied additional selection criteria. Paper  included only patients with instability of the coccyx, as determined using x-rays when sitting and standing. Paper  classified patients into four groups according to the cause of the pain.
In some studies the whole coccyx was removed, in others it was all removed if it was "mobile" and otherwise only the mobile part was removed. Studies  and  included some partial and some total coccygectomies. In study  some unsuccessful partial coccygectomies were later converted to total coccygectomies. Several papers referred to filing down any sharp edges on the end of the sacrum or the remaining coccyx vertebra to leave a smooth surface. In some studies all patients were given antibiotics to prevent infection.
Patients were interviewed between 1 and 26 years after their operations to assess how successful they had been. The criteria used for deciding the success of operations reported in these papers were subjective, based on the patients choosing between categories such as 'complete relief', 'substantial improvement', 'some improvement' or 'no improvement'. None were based on more objective criteria such as a comparison of tolerable sitting times before and after surgery. Subjective assessments are open to criticism, because the hopes of the patients for success and pressure from the doctors may bias results. However, in all these studies the assessments were carried out a year or more after surgery, when the effects of such pressures are likely to be small. Also many of these results were quite clear cut. It was clearest of all in the small study reported in Paper : `All patients were asked by telephone whether they could participate in a routine clinical and radiological follow up examination, but all refused because they were all symptomless, very satisfied and therefore believed that there was no need for such an examination.' The definitions of success used are shown in the table.
The results are shown in the table. The four different groups identified by reference  are listed as 3.1 - 3.4 in the table.
|Ref.||No. of patients||Additional selection criteria||Total/ partial removal||Success criterion||No. of successes||% success|
|1||28||Instability of coccyx||mobile part||relief 75- 100%||23||82|
|2||23||None||mobile part||complete sustained relief||21||91|
|3.1||25||Trauma||total||good or excellent relief||19||76|
|3.2||12||Spontaneous||total||good or excellent relief||7||58|
|3.3||7||Spine fusion||total||good or excellent relief||0||0|
|3.4||4||Childbirth||total||good or excellent relief||3||75|
|5||36||None||some of each||good or excellent relief||32||89|
|6||9||None||total||marked or complete relief||8||89|
|7||55||None||some of each||patient satisfied||46||84|
The results show that none of the seven patients whose pain was caused by an earlier operation to fuse parts of the spine (row 3.3) was helped by coccygectomy. Clearly the operation is not appropriate in these cases. If these seven are excluded, there remain 202 patients, of which 168 (83%) had a successful result. In group 3.2, who had coccyx pain of unknown origin, seven out of twelve (58%) had a successful outcome. This is a lower success rate than in other groups, though the numbers in this group are too small to draw definite conclusions.
The question of whether any patients are left in more pain after surgery than before is a very important one for patients. Many people in chronic pain with no other prospect of relief would be prepared to undergo an operation even if the success rate was fairly low, provided the risk of making the pain worse was negligible. Unfortunately, only 3 of the 7 papers specifically address this point.
Paper  gives no information, as its lowest category for the state of patients after surgery is 'less then 30% improvement'. Three papers [2,6,7] explicitly state whether or not any patients were worse off. These studies include 87 patients, two of whom (2.3%) were in more pain after surgery. Three other studies [3,4,5] have their lowest category for the patients condition after surgery as 'no improvement' or 'no relief', possibly implying that no patients had deteriorated. If these studies are included then two out of 181 patients (1.1%) were in worse pain after surgery.
Of course, pain can increase for other reasons, and for no obvious reason, so the deterioration in the condition of these patients is not necessarily due to surgery. Given the small numbers considered here, it is not possible to say more than that 1 or 2 % of patients may have increased pain after surgery. Hopefully, further research will clarify this.
Paper  reported 2 failures of the surgical stitches among 28 patients, but these did not pose any major problem. Paper  reported that healing was slightly delayed in one out of 23 patients. Paper  reported that four out of 60 patients had temporary urinary retention, two developed a hematoma (pocket of blood) which needed drainage, five developed a superficial infection and four a deep infection. Of the four with the deep infection only one had successful pain relief. Paper  reported one infection out of 10 patients which was treated with antibiotics. Paper  reported that 17 out of 36 patients (47%) reported some loss of sensation in the area of the coccyx, but none were troubled by this. None of the papers reported any loss of muscle control. Paper  reported that one out of 9 patients had the wound re-open on undertaking vigorous exercise three weeks after the operation. Paper  reported that four out of 55 patients developed superficial infections, one of which needed further surgery.
Overall it appears that the only complication reported in these papers which had a significant long-term effect in some cases was wound infection. Because of this, paper  recommends that antibiotics should be given to all patients as a preventive measure.
Paper  used an additional selection criterion, instability of the coccyx as determined from X-rays taken sitting and standing. The hope was that this would provide an accurate way of deciding who would benefit from surgery. It is disappointing to see that the success rate in this trial does not seem to be significantly different from that in other trials.
Paper  found that one third of coccyx pain patients had acute pain when moving from sitting to standing. All of these were found to have an unstable coccyx. However, not all people with an unstable coccyx had this acute pain on standing.
Paper  found that two out of 120 patients presenting with coccyx pain had cancer.
Paper  found that 36 out of 120 patients had a bony prominence on their coccyx.
Paper  found that total coccygectomy gave better results than partial coccygectomy, and recommends the use of total coccygectomy. Paper  also reported results for partial or total removals, again with better results for total coccygectomy, though in this case the numbers are too small to draw definite conclusions. However, in the studies reported in papers  and , partial coccygectomy (removal of only the part which is mobile) gave good results. Six patients in study  who had a partial removal with poor results were then given a total coccygectomy, and four improved as a result.
Several papers reported that the average time for coccyx pain to fade after surgery was 3-5 months, but could be more than a year. (Note that Linsey found that in her case it took 2 years.)
Paper  reported that none of the 51 coccyxes examined had a fracture, and none had a complete dislocation. The paper suggests that these conditions are rarely involved in coccyx pain. Complete dislocation, though rare, is not unknown - Rory Greenwell's surgeon told him that his coccyx was floating unconnected. However, many patients are told after X-ray that they have a fracture, in conflict with the finding in paper  after surgery. I wonder if this could be because most medical books wrongly say that the coccyx is normally fused into one piece, and when doctors see it on the X-ray in 2 or 3 pieces (which is actually more common) they think it must be fractured. Note that Susan was told at first that she had two fractures, then after a second X-ray and bone scan was told that there was no fracture, and fractures were difficult to diagnose in this area.