1–year follow–up of 920 hip and knee arthroplasty patients after implementing fast–track

Acta Orthop 2015; 86: 78-85. Early online. DOI 10.3109/17453674.2014.957089

A calculation error has been made concerning the two questions (1) “How does the leg that was operated on work today compared to before surgery?”, and (2) “Based on your experience to date, would you go through the surgery again?”. The results were calculated based on all patients included in the study, and not only the patients attending the 1 year follow-up who were asked these specific questions. The results are now corrected and marked with red in the following page, which should replace the fourth page in our article.

(4.3–7.0). The revision rate was 2.9% (2.0–3.9), of which 1.6% (0.9–2.3) was caused by infections and 1.0% (0.4–1.5) by dislocations. At 1–year follow–up, 93% reported improved functionality in the operated limb and 95% reported that they would have been willing to have the surgery all over again.

Figure 2. Pain at rest (upper panel) and mobilization (lower panel) at 8 time points, from preoperatively until 1 year postoperatively. Lines represent mean pain score for each patient group.

For TKA patients who underwent primary surgery, 96% were mobilized in the recovery unit and the mean time from surgery to mobilization was 3.2 (1.4) h. Mean LOS was 3.1 (0.8) days. 82% of the patients were discharged directly to their homes. Total re–admission rate within 1 year was 10.1% (8.4–11.8). The revision rate was 3.3% (2.3–4.3), of which 1.4% (0.7–2.0) was caused by infections and 1.4% (0.7–2.0) by mechanical causes. At 1–year follow–up, 85% reported improved functionality in the operated limb and 86% reported that they would have been willing to have the surgery all over again.

Revision surgery

For THA patients who underwent revision surgery, 91% were mobilized in the recovery unit and the mean time from surgery to mobilization was 4.6 (2.2) h. Mean LOS was 4.2 (1.6) days and 44% of the patients were discharged directly to their homes. Total re–admission rate within 1 year was 5.6% (4.3–6.9), all of which was caused by infections. At 1–year follow–up, 66% reported improved functionality in the operated limb and 72% reported that they would have been willing to have the surgery all over again.

For TKA patients who underwent revision surgery, 89% were mobilized in the recovery unit and the mean time from surgery to mobilization was 4.9 (1.9) h. Mean LOS was 3.9 (2.2) days and 67% of the patients were discharged directly to their homes. Total re–admission rate within 1 year was 7.1% (5.7–8.6), of which 3.6% (2.5–4.6) was caused by infection and 3.6% (2.5–4.6) was due to mechanical causes. 1 year postoperatively, 68% reported improved functionality in the operated limb and 74% reported that they would have been willing to have the surgery all over again.

Table 2. Patient–reported outcome scores (PROMS) in the 4 different groups: EQ–5D, HOOS–PS, KOOS–PS, and the physical function scores HHS and KSS. For abbreviations, see text. All scores were obtained preoperatively, 8 and 12 weeks postoperatively, and 1 year postoperatively. Data are mean (SD)

	Primary surgery		Revision surgery
	THA	TKA	THA	TKA
Preoperatively
EQ–5D index score	0.42 (0.30)	0.49 (0.29)	0.42 (0.32)	0.38 (0.36)
HOOS–PS/KOOS–PS	45 (16)	47 (13)	49 (18)	49 (11)
HHS	53 (14)		52 (24)
KSS knee score		41 (14)		49 (19)
KSS function score		56 (22)		43 (23)
8/12 weeks postoperatively
EQ–5D index score	0.76 (0.21)	0.68 (0.24)	0.62 (0.32)	0.53 (0.32)
HOOS–PS/KOOS–PS	22 (14)	37 (11)	33 (16)	41 (12)
HHS	82 (13)		69 (20)
KSS knee score		61 (19)		64 (15)
KSS function score		59 (25)		45 (28)
1 year postoperatively
EQ–5D index score	0.79 (0.24)	0.76 (0.230)	0.67 (0.27)	0.62 (0.25)
HOOS–PS/KOOS–PS	17 (15)	30 (15)	27 (15)	34 (15)
HHS	89 (14)		77 (20)
KSS knee score		75 (19)		62 (13)
KSS function score		78 (24)		65 (25)

Discussion

After implementation of the fast–track course, the number of weekly hip and knee arthroplasty surgeries was increased from 7 to 17. Patient satisfaction was high in all parts of the treatment chain, with a mean score of 9.3 (1.2) out of a maximum of 10. 95% of hip patients and 86% of knee patients were satisfied with the results 1 year after the operation. In addition, LOS was reduced by approximately 5 days for both hip patients and knee patients. The satisfaction with the LOS had a mean score of 8.9 (1.7) out of a maximum of 10. These findings are in line with other publications concerning fast–track arthroplasty (Husted et al. 2008, 2010a). Hospital logistics and clinical features are crucial for the LOS, and reduced LOS reduces costs without compromising treatment quality (Husted et al. 2008, 2010d, 2012). The numbers of re–admissions and revisions in the present study were lower than those previously reported (Husted et al. 2008, 2010b). These results demonstrate that even though the treatment is more effective, as indicated by the increased number of patients operated annually and with a reduced LOS, it does not compromise patient satisfaction or treatment quality. Fast–track also provided good results in non–septic revision surgery, as has been reported previously by others (Schneider et al. 2009, Husted et al. 2011b, Jorgensen et al. 2013).

Measurement of pain with an NRS is a practical method to use, since it is easy to understand and the patient does not need clear vision or paper and pen to describe the degree of pain—in contrast to the visual analog scale (VAS). An NRS of 3 or less corresponds to “mild pain” on the VAS, whereas 4–6 corresponds to “moderate pain” and 7–10 to “severe pain” (Breivik et al. 2008). Pain, dizziness, and general weakness are main causes of prolonged postoperative hospitalization. Pain is also a limiting factor for early postoperative mobilization and physical activity (Husted et al. 2011a). In general, the patients in our study had only mild pain directly after or during the evening following surgery. We found that the maximum pain experienced was during rest in the evening after surgery, both for primary hip and knee patients and for revision hip and knee patients (2.9 (2.0), 2.9 (2.0), 2.3 (1.8), and 3.9 (2.6), respectively). From day 2 postoperatively to patient discharge, the pain both at rest and during mobilization was reduced and sustained at the “mild pain” level. These results suggest that pain was no limitation regarding early discharge.

For optimal rehabilitation and assured function, patients need continuous analgesic treatment after discharge (Andersen et al. 2009, Husted et al. 2011a). Holm et al. (2010) found that pain has a limited influence on functional recovery after the first postoperative day after TKA, thereby allowing early physiotherapy. They found that 90% of the patients were mobilized after the first postoperative day, which agrees with our findings. We found that over 90% of the primary patients and about 90% of the revision patients were mobilized at the recovery unit, and the mean time from surgery until mobilization was about 3.5 h and 4.5 h, respectively. Patients who were not mobilized were either immobilized due to perioperative complications, illness, or inability to cooperate.

Early mobilization is important to reduce the risk of thrombosis and to initiate rapid recovery (Husted et al. 2008, 2010b,c). The numbers of deep–vein thrombosis and pulmonary embolism could be underreported in our study. These diagnoses had to be verified by ultrasonography or CT before being reported to the registry.

All the patients had to fulfill the fast–track discharge criteria before leaving hospital, either to their homes or to the rehabilitation institution. Before the fast–track implementation, there were no standardized discharge criteria, and the patients usually went to a rehabilitation institution. Approximately 80% of the primary patients in our study were discharged directly to their homes, which is a lower percentage than can be found in publications from Denmark (Husted et al. 2008). In Norway, it has been a tradition to discharge patients to rehabilitation institutions depending, for example, on the patient’s housing conditions, care facilities, comorbidities, and distance from hospital. Now that fast–track has been implemented, this trend is about to shift. Only 44% of the revision patients were discharged to their homes due to their reduced functionality postoperatively—the result of greater surgical trauma, as full synovectomi and intramedular reaming for femoral and tibial components are used in the revision surgeries. At postoperative follow–ups for all groups at week 8 or 12, and at 1 year, the mean pain score was relatively low, as expected, with values of less than 3 on the NRS. These findings are in line with the results of other studies (Rostlund and Kehlet 2007, Holm et al. 2010), demonstrating low levels of postoperative pain after fast–track joint replacement surgery.

Even though the main purpose of THA and TKA surgery is pain relief, regaining HRQoL and functionality is considered to be the ultimate goal after joint replacement (Woolf and Akesson 2001). Thus, to evaluate the treatment outcome, it is important to measure outcomes from the patient’s perspective. In a fast–track treatment set–up without any formal intensive rehabilitation after discharge, it was found that THA and TKA patients at 1–year follow–up had regained health similar to that of an age– and sex–matched population in Denmark when measured by HRQoL (EQ–5D) (Larsen et al. 2010, 2012). We found similar improvements (using EQ–5D scores) one year after primary THA and TKA to those reported by Larsen et al. However, the patients in our study did not reach the same HRQoL levels 1 year postoperatively, which may have been due to the different demographics of the patient cohorts. Patients included in the hip study (Larsen et al. 2010) had a diagnosis of unilateral primary arthritis, and in the knee study (Larsen et al. 2012) patients with unicompartmental knee arthroplasty were included. In the present study, patients with bilateral hip arthroplasty were included and patients with unicompartmental knee arthroplasty were excluded. Thus, our patients may have been affected by the disease to a greater extent than the patients in the Larsen study, therefore resulting in both higher preoperative and 1–year HRQoL scores. EQ–5D results from the Swedish Arthroplasty Register showed that THA patients had mean scores of 0.42 preoperatively and 0.78 after 1 year (Rolfson et al. 2011), which is similar to what we found. A cohort study from Karolinska University Hospital in Sweden (Jansson and Granath 2011) also found similar EQ–5D results in THA patients pre- and 1 year postoperatively (0.49–0.80, and TKA patients pre- and 1 year postoperatively (0.51–0.73).

The THA patients in the study by Larsen et al. (2010) did not regain the same level of health as the age– and sex–matched normal population, which was based on the disease–specific HHS outcome one year postoperatively (88 vs. 94). However, the results were similar to our findings (HHS = 89) 1 year postoperatively. Any HHS score within these values is close to “excellent” (i.e. close to > 90) on the HHS grading scale, which indicates that even though THA patients have good functionality at 1 year postoperatively in disease–specific terms, they do not reach the level of the normal population. In another study (Medalla et al. 2009), the postoperative KSS outcomes, knee score (81) and physical function score (71), were similar to the postoperative scores that we found (75 and 78, respectively). However, patients in the Medalla study gave their postoperative scores 2 years after surgery, as compared to 1 year in the present study. Also, the patients in the Medalla study generally had a higher knee score and a lower physical function score than in our study. This might be explained by the different postoperative testing times, since knee function is directly related to the surgery and physical function is related to the physical fitness of the patient, which is not directly influenced by the surgery. Scores between 70 and 80 correspond to a “good” knee and functionality score on the KSS grading scale. The knee and physical function scores in the present study were both > 75, which is close to “excellent” (> 80), indicating good functionality one year postoperatively.

In a study from Toronto, Canada (Davis et al. 2009), based on 201 THA and 248 TKA patients, the improvements in HOOS–PS (55%) and KOOS–PS (33%) from preoperatively to 6 months after surgery were similar to what we found. This similarity was apparent in HOOS–PS from before surgery to 3 and 12 months after surgery (51% and 62%, respectively), and in KOOS–PS from before surgery to 2 and 12 months after surgery (21% and 35%), where the patients in both studies reported less difficulty in performing daily activities postoperatively. In light of our findings and those previously reported by others, fast–track is a treatment that gives good postoperative results, both in generic and disease–specific terms.

The overall loss of participants at follow–up in our study 1 year postoperatively was 14%, which is less than that reported by others (26%) (Larsen et al. 2012). A higher proportion of revision patients than primary patients were lost to follow–up at 1 year. Local hospitals often refer patients scheduled for revision surgery to the university hospital. The follow–ups are often done at the local hospital, which could explain the higher loss of patients observed in this group. Patients eligible for follow–up had different reasons for not turning up. In some cases the secretary had not scheduled an appointment, some patients had forgotten the appointment, some patients did not want the consultation, and patients with severe complications at the first follow–up were excluded from the second follow–up. However, all patients with complications resulting in revision were followed up, as they were re–admitted to the same university hospital.

In summary, we found improved efficiency, high patient satisfaction, and low revision rates together with improved health–related quality of life and functionality with the fast–track course implemented in a Norwegian university hospital, indicating that fast–track is a favorable and feasible method of treatment for both primary and revision THA and TKA patients.

All the authors contributed to the literature search, study design, data collection, data analysis, and interpretation. All the authors contributed substantially to drafting of the article and they also revised it for important intellectual content.

No competing interests declared.