Fever is a fairly common symptom in the first few days after various types of surgery, including joint replacement surgery. At present, fever following joint replacement is thought to be caused by: ① non-infectious factors, such as inflammatory reactions generated by different cytokines in human body(9, 10); ② Infectious factors such as incision infection, periprosthetic infection, urinary/ respiratory tract infection and so on(11). Above all, the surgeon's primary concern is that a deep infection of the wound or prosthesis might occur. Because PJI is a catastrophic complication that might occur after total joint arthroplasty. However, our concern often leads to unnecessary diagnostic examinations such as chest radiographs, urine analysis, urine and blood cultures(3, 6, 12–14). These examinations seldom create useful diagnostic value but will increase the financial burden of patients.
The quantity of publications focused on pyrexia following TJI is quite limited, with some different conclusions. Athanassious et al(8) studied on 195 TKA patients and 146 THA patients, respectively. What they found was a 36% fever occurrence among patients. In addition, POD 1 was the day when a patient would most likely develop a fever, and POD 2 was the day when a maximum temperature would be measured on a patient body. They believed that fever was a normal phenomenon and should not cause a serious concern in the immediate postoperative period in most patients. It might be unnecessary to conduct urinalysis or urine culture in the evaluation of a fever in the immediate postoperative period. Chest x-ray would not become necessary unless the patient had suffered multiple fever attacks during his/her hospital stay.
Lu et al(5) researched on 579 TKA patients and 383 THA patients. They found that fever occurred in 48.2% of patients (≥ 38℃), most commonly on POD1, and the maximum body temperature also most commonly appeared on POD1. Besides, they drew a conclusion that consecutive days of fever and fever ≥ 39℃ had something to do with infection in patients after TJA.
Kennedy et al(15) investigated 90 TKA patients and discovered that 17% of patients experienced fever (≥ 39℃), most commonly on POD1. Their findings were: There were significant independent correlations between high pyrexia and both blood transfusion and fall in haematocrit; Pyrexia and the presence of an infective focus were irrelevant. And they concluded that pyrexia occurring in the first few days following TKR was a normal physiological process. A recent study investigated 168 TKA patients and reached a conclusion that the temperature rose most remarkably on postoperative days 1 and 2, but only 4 patients caught fever up to 39 ℃. This study also showed an negative correlation between hemoglobin loss and postoperative body temperature(16).
Ghosh et al(7) conducted a research on 170 TKA patients to find that patients’ fever occurrence rate was 36.5% (≥ 38℃), and that there was no relation between pyrexia and infection, allogenic blood transfusion, haemoglobin loss, use of urinary catheter, rheumatoid arthritis, anesthetic type, or previous pyrexia following TKA. Pyrexia in the first 5 days after TKR was usually a normal physiological response and excessive concern about the presence of infection would be unnecessary.
Shaw and Chung(17) examined 100 patients who had undergone TKA and another 100 patients total hip replacement (THA). They found that the average maximum temperature peaked on the POD1 and that 19% had a fever of 39°C or above at some point, while achieved no finding of correlation between blood loss or transfusion and pyrexia was found. Therefore their final conclusion was that post-operative fever after TJA was a normal physiological response and that a full work-up for fever alone was not justified.
This study shows that only 4% of patients experienced a fever after TJA (3% of patients experienced a fever after TKA and 6% of patients experienced a fever after THA). In contrast, the prevalence of fever in the current study is far lower than previous studies. We believe that low pyrexia rate may be related to the widely used tranexamic acid or other hemostatic drugs. Besides, the patients in this study were treated on PODs 2 to 4 with NSAIDs for pain inhibition. This may also be a potential cause for low fever rates.
In addition, this study reveals that the first fever after TKA was most common on POD2, and the first fever after THA was most common on POD1. The maximum body temperature was most common on POD2 in patients with fever after TKA, and POD1 in patients with fever after THA. For each group, the first fever and the maximum body temperature occurred on the same day, which is consistent with study reported by Tai et al(16). in patients after TKA. The timing of maximal body temperature and first fever is likely due to surgical trauma and secondary systemic inflammatory responses related, which decreases with time as reflected by body temperature reduce.
We found no significant correlation between the development of a fever and hypertension, diabetes, blood transfusion, use of urinary catheter, an aesthetic type, surgical time, postoperative drainage volume and pre-operative albumin. This is consistent with the conclusion of Andres et al(10). While there was a significant correlation between fall in hemoglobin and the development of a fever in patients after THA, but it did not occur in patients after TKA. This is consistent with the conclusion of Ghosh et al(7, 16, 18) in patients after TKA. The difference may be related to the use of tourniquet, because many studies had indicated that tourniquet application could reduce surgical time, intraoperative blood loss and total blood loss.
Furthermore, we found that there was significant association between the development of a fever and post-operative CRP. Surgical stress, postoperative infection and tissue injury can increase the level of serum C-reactive protein(19). The smaller the surgical trauma was, the lower the blood C-reactive protein level would be. There was a positive correlation between the increase of body temperature and the increase of C-reactive protein after joint replacement. C-reactive protein is an important marker of inflammatory response, which is synthesized by hepatocytes. Its response often precedes clinical symptoms, including fever(20).
Limitations of this Study:
There are some limitations to our study. First of all, this is a retrospective analysis. Second, all the patients in this study were treated with non-steroidal anti-inflammatory drugs for analgesia, which might have alleviated the fever response of the patients. Third, all patients received antibiotics on POD0 ~ 2, which might also have reduced the febrile response. Fourth, this study did not explore the relationship between postoperative fever and infection, and we suggest that a prospective study is needed to verify our results.