In the current study, the 90-day unplanned readmission rate after surgical treatment for TSS is 1.7%, the causative factors include CSF cyst and poor incision healing followed CSF leakage, intraspinal hematoma, wound dehiscence, surgical site infection, inadequate decompression. The peak period of readmission occurred from 10 to 40 days after surgery. When compared to the non-readmitted patients, diagnosis of OPLL + OFL, circumferential decompression, dural injury, long hospital stay is more to be seen in readmitted patients, we suppose that these four factors are not independent but interrelated. Patients with OPLL combined with OFL always need circumferential decompression, which poses high risk of dural injury and increases the hospital stay subsequently.
It is unavoidable to injure the dura sac in patients with OLF and OPLL due to the adhesions between ossified ligaments and the dura sac. Moreover, ossification of dura sac could increase the technical difficulty of surgical decompression and the risk of postoperative CSF leakage [20]. Previous studies demonstrated that CSF leakage did not affect the long-term clinical outcomes, but we find that cerebrospinal fluid cyst and poor incision healing followed CSF leakage could increase the risk of unplanned hospital readmission after surgical treatment of TSS in the short term (within 90 days of the first surgery) [21, 22]. The main complaint of patients with cerebrospinal fluid cyst is local bulge under the incision, headache and lower extremities weakness, and they admitted from the 1 month to 3 months after surgery, while patients with poor incision healing followed CSF leakage always readmitted within first month after surgery. For reducing the incidence of postoperative CSF leakage and the rate of unplanned admission, we suggest four possible surgical strategies. Firstly, for patients that are at high risk of dural tear in the preoperative evaluation, such as dura ossification, floating method should be recommended as the primary surgical method, instead of slitting dura procedure. Secondly, try to prevent dural tear by improving surgical instrumentation, it has been reported that the decompression surgery could be completed safely and quickly when use of ultrasonic bone curette in posterior thoracic decompression [23]. Thirdly, try to repair dural tear directly if possible using fibrin glue, gelfoam and artificial dura, and anaesthetists should ask the patient to perform a few Valsalva manoeuvres to confirm absence of CSF leakage. Fourthly, it is important to ensure tight suturing of the muscle and fascia layers [24]. Postoperatively, bedrest and compressive dressing after the removal of drainage are also necessary during the hospitalization [25].
Postoperative spinal epidural hematoma (PSEH) is a rare complication after spinal surgery with the incidence ranged from 0.1–0.4%, often results from incomplete intraoperative haemostasis, blocked drainage tubes or coagulopathy, it can lead to devastating neurological deficits including sensory disturbance, lower extremities weakness, and bowel dysfunction [26].PSEH often develops in a few days, especially on the day of surgery. However, on rare occasions, PSEH can occur more than three days, up to two weeks after the initial surgery. Uribe J et.al. defined it as the delayed PSEH, and reported the incidence was 0.17% [27]. In the current study, a female patient experienced circumferential decompression for T3-4 OPLL combined with OFL, she discharged on postoperative day 5, while complained lower extremities weakness on day 7, then readmitted and diagnosed as delayed PSEH. We cannot make any conclusion based on only one patient data, but it should be clear that any attempt to prevent PSEH may potentially decrease the incidence of unplanned admission in surgically treated patients with TSS.
Wound dehiscence is uncommon in spine surgery, but it could result in unplanned readmission if occurred. Winward Ch et al. reported that complications of SSI and wound dehiscence were most common readmission reasons, accounting for 15.6% of all readmission following posterior cervical fusion [28]. Chibuikem Ak et al. also find that SSI and wound complications were the most common causes of readmission after surgical treatment of common lumbar pathologies [5]. In the current study, only two patients presented readmission due to wound dehiscence in surgically treated patients with TSS, it is not the primary contributor, differing from the cervical and lumbar spine. Anatomically, the incision in posterior thoracic decompression surgery results in greater susceptibility to tension from paraspinal muscles due to the thoracic kyphosis, pedicle screws would further increase the tension, especially in patients with long decompression level through kyphosis apex. Consequently, there is a greater stress on the healing skin and fascial layers, finally lead to wound dehiscence. Ando K, et al. have proved that dekyphosis using multilevel Ponte osteotomies could provide indirect decompression of the spinal cord [29]. We suggest that thoracic kyphosis decrease may also prevent wound dehiscence by reducing the tension from paraspinal muscles and may decrease the incidence of unplanned readmission in surgically treated patients with TSS.
SSI is a relatively common complication after spinal surgery, with reported incidence ranges from 0.1% to 10.9% [16]. SSI increases the morbidity, mortality, length of hospital stay, readmission, and health care costs. Two patients presented unplanned readmission due to SSI in the current study, proving it again that try to prevent SSI may decrease the rate of readmission. It has been proved that risk factors of SSI include age, ASA score, obesity, diabetes, smoking, radiation therapy, psoriasis, chronic skin conditions, use of instrumentation, bone graft harvesting [30]. The identified predictors of SSI can improve identification of high risk patients and provide a key target for intervention.
Inadequate original surgical decompression is rarely reported in the previous literature, there is a paucity of published data on the incidence of patients with myelopathy who have undergone a surgical decompression for TSS that was subsequently found to be inadequate. This lack of data may be due to the postoperative CT is not commonly obtained unless, or until, a subsequent neurologic decline is noted. Definitely, revision surgery may be required for most of the patients that with inadequate original surgical decompression, then lead the unplanned admission. In the current study, two OLF patients experienced unsatisfactory symptom relief and functional improvement at three months follow up, then confirmed inadequate original surgical decompression through CT. We suppose this type of readmission should be classified as iatrogenic, because it is possibly derived from the improper surgical plan by surgeons, more careful should be implemented in the preoperative surgical planning, one more segment decompression both at caudal and cephalad could decrease the risk of inadequate original surgical decompression.
This retrospective study has several limitations. First, risk factors for readmission were not assessed through Logistic regression analysis because of the relatively small sample size. Second, this is a single institution study, the conclusion may be not necessarily generalizable or representative of the population at large, but it could provide the advantage of consistency in surgical practice and clinical care.