Abstract
Purpose
The aim of this study was to examine factors associated with a prolonged length of stay (LOS) in patients over 75 undergoing a total knee replacement (TKR).
Methods
Patients over 75 undergoing a TKR at our institution from January 2008 to February 2009 were identified (n = 112). Patient and operative factors previously shown to affect length of stay were identified. Patient notes were reviewed for details on each of these and data analysed for their effect on length of stay. Discrete data were analysed for their effect on post-operative length of stay using either the Mann-Whitney U test or the Kruskall-Wallis test and continuous data analysed with the Spearman’s rank correlation coefficient.
Results
The following factors were associated with length of stay at the 95 % confidence level: patient age, pre-operative mobility and the use of walking aids, BMI, whether the patient was able to mobilise within 24 or 48 hours of the surgery, the day on which the patient first walked ten metres and achieved 90° active knee flexion, pre and post-operative haemoglobin and the need for a blood transfusion.
Conclusions
Pre-operative use of walking aids, peri-operative haemoglobin concentration, failure to mobilise early following the operation and post-operative complications (including the need for a blood transfusion) seem to be the significant factors associated with a prolonged stay in hospital in the over 75 year olds.
Introduction
Osteoarthritis (OA) is the most common disease affecting joints around the world. Epidemiological studies show that there is radiological evidence of osteoarthritis in the majority of the people over 65 and in about 80 % of those over 75 [1]. With an aging population and an increasing lifespan, it is not difficult to imagine that the financial burden of OA on health services is likely to increase over the next few years. In the financial year 08/09, over 60,900 primary total knee replacements (TKRs) were performed in England alone and the trend over the last few years suggests this is likely to increase further [2]. With the current financial climate, the UK government is pledging to deliver £15–20 billion in savings in the health sector by 2013–14 [3]. This has resulted in an emphasis on streamlining services and in particular on reducing the length of stay (LOS) in hospital following surgical procedures. Indeed, reducing the LOS has been shown to have benefits that go beyond the economic benefits. Elderly patients are most prone to post-operative complications and in many cases are the candidates for major surgery such as joint replacement. Studies have also shown that this group of patients are at a greater risk of mortality associated with surgical site infection [4].
Following the recent introduction of an Integrated Care Pathway (ICP) for patients undergoing a total hip or knee replacement at our institution, the target for length of stay following a TKR was set at five days. There have been a number of studies that have addressed the possible factors affecting length of stay and outcomes following TKR, with some variability in the findings [5–7]. However, there has been no study specifically pertaining to the over 75 cohort. In view of the aging population and increasing willingness on the part of surgeons to operate on older patients, we feel that specific investigation into older populations is warranted. The aims of this study therefore were to investigate the factors that influence post-operative LOS in patients aged over 75, and how this compares with those aged 75 and under.
Patients and methods
Sample
A retrospective cohort analysis of 387 patients who underwent a primary TKR at our institution between January 2008 and February 2009 was carried out. Of these, 112 patients were over the age of 75 at the time of the procedure. Table 1 describes the characteristics of the sample. Medications including warfarin, anti-platelet therapy and non-steroidal anti-inflammatory agents that patients had been on were stopped pre-operatively. All TKRs were performed with a high thigh tourniquet and all the patients received post-operative thromboprophylaxis in the form of enoxaparin and TED stockings. The exclusion criteria were: revision knee replacement, simultaneous other joint replacement, medically unrelated confounding factors (e.g. coincidental diagnosis of a brain tumour in the post-operative period). Patients were included regardless of the indication for primary TKR (osteoarthritis, rheumatoid arthritis or other inflammatory arthropathy) or pre-operative comorbidity. Exclusion criteria were lenient in comparison with other studies [8] because the authors wanted to assess the factors that affect LOS in normal clinical practice, rather than in a trial situation. Only two patients were excluded: one underwent a simultaneous bilateral TKR and the other was coincidentally diagnosed with a brain tumour in the post-op period.
Table 1.
Pre-operative characteristics of the sample
| Characteristics | Values |
|---|---|
| Age | Mean: 80.7 |
| Range: 75–93 | |
| Sex | 54 males (48 %) |
| 58 females (52 %) | |
| Marital status | 4 % single |
| 57 % married | |
| 35 % widowed | |
| 4 % divorced | |
| Pre-operative mobility | 35 % no walking aid |
| 53 % walking stick(s) | |
| 12 % crutch(es), walking frame or wheelchair | |
| ASA grade | Grade 1: 12 % |
| Grade 2: 64 % | |
| Grade 3: 22 % | |
| Grade 4: 2 % |
Measurements
Length of stay was calculated as the number of days in hospital from the day of surgery to the day of discharge, with day of surgery being day 0. At our institution, patients are deemed fit for discharge if there is a consensual agreement between members of the multidisciplinary team (including the consultant, physiotherapist and occupational therapist) that the patient is medically stable and functional ability is sufficient to allow discharge to their home environment. All patients were discharged to their own pre-operative living quarters. While the ability to flex the knee to 90° was desirable at discharge, it was not a formal discharge criterion.
Thirty-two demographic, pre-operative, intra-operative and post-operative factors were identified from the literature as having been previously investigated in relation to post-operative LOS following a TKR (Table 2). Patient notes were thoroughly examined for all the details on these factors and a database created. For each factor, the number of data sets available is indicated in Table 2. All the data was maintained on a Microsoft Excel spreadsheet (Microsoft Corp, Redmond, Washington). Comorbidities were weighted according to the Charlson score [9]. The comorbidities were summed and patient awarded a final comorbidity score.
Table 2.
Significance of patient, pre-, intra- and post-operative factors for length of stay
| Factors | Available data | Statistical test | P | R |
|---|---|---|---|---|
| Patient age | 112 | Spearman’s | 0.0138 | 0.0215 |
| Patient sex | 112 | MWU | 0.075 | |
| Marital status | 112 | KW | 0.170 | |
| Lives with (alone, live with someone) | 112 | MWU | 0.045 | |
| Support available on discharge | 109 | MWU | 0.832 | |
| Pre-op mobility (independent without aids, walking stick, crutches, frame, wheelchair) | 109 | KW | 0.00941 | |
| Comorbidity score | 112 | Spearman’s | 0.137 | 0.140 |
| BMI | 112 | Spearman’s | 0.00839 | −0.246 |
| ASA grade | 112 | KW | 0.400 | |
| Other previous joint replacements | 112 | KW | 0.422 | |
| Pre-op Hb | 112 | Spearman’s | 0.000485 | −0.301 |
| Pre-op platelets | 112 | Spearman’s | 0.186 | 0.117 |
| Type of anaesthetic (general vs. spinal) | 112 | MWU | 0.84 | |
| Use of a local nerve block | 112 | KW | 0.395 | |
| Tourniquet time | 109 | Spearman’s | 0.059 | −0.181 |
| Weekday of surgery | 112 | KW | 0.142 | |
| am vs. pm surgery | 112 | MWU | 0.270 | |
| Post-op Hb | 112 | Spearman’s | 0.000886 | −0.287 |
| Post-op platelets | 112 | Spearman’s | 0.560 | 0.052 |
| Hb drop | 112 | Spearman’s | 0.560 | 0.051 |
| Platelet drop | 112 | Spearman’s | 0.478 | 0.063 |
| Need for post-op blood transfusion | 112 | MWU | 1.15E−05 | |
| Volume of autologous blood reinfusion | 108 | Spearman’s | 0.915 | −0.010 |
| ITU/HDU admission | 112 | KW | 0.0597 | |
| Day first shown exercises by the physiotherapist | 112 | KW | 0.269 | |
| Able to mobilise within 24 hours of surgery | 112 | MWU | 0.0140 | |
| Able to mobilise within 48 hours of surgery | 112 | MWU | 0.00200 | |
| Post-op complications | 110 | KW | 5.40E−09 | |
| Use of oral strong opioid analgesia | 112 | MWU | 0.00318 | |
| Mobility on discharge (discharged on crutches or a walking frame) | 112 | MWU | 0.00173 | |
| Post-op day mobilised 10 m | 112 | Spearman’s | 2.26E−06 | 0.420 |
| Post-op day achieved 90° knee flexion | 92 | Spearman’s | 1.34E−13 | 0.677 |
BMI body mass index, MWU Mann-Whitney U test, KW Kruskall-Wallis test, Spearman’s Spearman’s rank correlation coefficient
Statistical methods
SPSS statistics (version 17.0, SPSS Inc., Chicago, Illinois) was used for all statistical analyses. The correlation between continuous data (age, comorbidity score, BMI, pre-operative Hb, pre-operative platelet count, tourniquet time, first post-operative Hb and platelet count, Hb drop and platelet count drop, volume of autologous blood reinfused, post-operative day on which the patient was able to achieve 90° of active knee flexion and able to walk ten metres) and post-operative LOS were analysed using Spearman’s rank correlation coefficient as the variables were not found to be normally distributed in a Shapiro-Wilks test.
Discrete data were analysed for their effect on post-operative LOS using either the Mann-Whitney U test (variable had two categories) or the Kruskall-Wallis test (variables with greater than two categories). All tests were two-tailed. Results were considered statistically significant at the 95 % confidence level and highly statistically significant at the 99.9 % confidence level (p < 0.001).
Results
The mean LOS was 6.25 days with a range from one to 37 days. Twenty-five patients stayed ten to 20 days. Only one patient had a LOS greater than 20 days. This patient was 80 years old and had a LOS of 37 days. Their hospital admission had been complicated by slow post-operative mobilisation, a urinary tract infection and development of new onset left bundle branch block subsequently treated as acute coronary syndrome under the guidance of the cardiology department. One patient had a LOS of one day. This patient was a 75-year-old male with excellent pre-operative mobility. He had first mobilised on the day of operation, suffered no complications and was discharged on the first post-operative day.
There was no statistically significant association between the following and LOS: sex, marital status, the availability of support at home on discharge and interestingly, comorbidity score or ASA grade. There was no significant difference between patients operated upon on different days of the week or those operated upon in the morning compared to the afternoon. It was also found that the type of anaesthetic (spinal or general anaesthetic), the use of a femoral or combined femoral and sciatic nerve block and tourniquet time were not associated with LOS. Thirty-nine patients received a femoral or combined femoral/sciatic nerve block.
The following factors were found to have a statistically significant association with the LOS: pre and post-operative haemoglobin (Figs. 1 and 2) and the need for a blood transfusion, patient age, pre-operative mobility and the use of walking aids (Fig. 3), BMI, whether the patient was able to mobilise within 24 or 48 hours of the surgery, the day on which the patient first mobilised 10 m (Fig. 4) and achieved 90° active knee flexion (Fig. 5). Interestingly, the post-operative drop in haemoglobin was not statistically significant. There were no bleeding complications secondary to use of thromboprophylaxis.
Fig. 1.
The effect of pre-op Hb on mean length of stay (LOS)
Fig. 2.
The effect of post-op Hb on mean length of stay (LOS)
Fig. 3.
The effect of pre-op use of any walking aid on post-op length of stay (LOS)
Fig. 4.
Percent of patients versus the post-operative day that they were able to achieve mobilisation to 10 m
Fig. 5.
Percent of patients versus the post-operative day that they were able to achieve knee flexion to 90 degrees
A poorer pre-operative mobility status was found to be associated with a greater LOS, with those independent with no walking aids staying a mean of 4.89 days, compared with 6.14 days in those using a walking stick and 6.85 days in those having to use crutches, a walking frame or a wheelchair.
The results have shown that early post-operative mobilisation, as measured by the earliest day on which the patient was able to walk ten metres and achieve 90° of active knee flexion, is associated with a reduced LOS. Patients who were able to walk ten metres or achieve 90° active flexion on the first post-operative day had a mean LOS of 4.31 and 3.50, respectively. In comparison, those who walked ten metres on the third day had a mean LOS of 5.71 days, and those who only achieved 90° knee flexion on day 3 had a mean LOS of 5.38 days.
Discussion
Pre-operative and patient demographic factors that were found to be associated with increased LOS were increasing patient age, poorer pre-operative mobility and use of walking aids, and lower BMI. It is well-documented that obese patients undergoing TKRs are at greater risk of prosthetic infection [10] and wound complications; the association between lower BMI and increased LOS is not one that has been corroborated in previous studies. It could be explained here by the fact that slimmer and frailer patients, especially in the elderly population, may struggle with rapid mobilisation post-operatively and hence this may hinder their recovery, resulting in an associated increased length of stay. There are, however, studies that have demonstrated no relation between BMI and LOS [5].
The association between poorer pre-operative mobility (i.e. walking unaided compared with walking with an aid or need for a wheelchair) and increased LOS is one that has been replicated in previous studies [5]. This could be explained in terms of differing gait patterns and increased muscular strength in those walking unaided and may suggest a role for aggressive pre-operative physiotherapy to improve the patients’ mobility status prior to undergoing TKR.
The results have shown that early post-operative mobilisation is associated with a reduced LOS. Studies have shown that flexion rehabilitation regimes show a better outcome than extension regimes, with patients not only being discharged earlier, but also showing greater range of movement at six-week follow up [11]. The results highlight the importance of early and aggressive post-operative mobilisation and physiotherapy. The benefits of early mobilisation go beyond mobility, as early post-operative mobilisation has been shown to reduce the risk of venous thromboembolism as well [12]. With the increasing popularity of integrated care pathways for patients undergoing TKRs, joint replacement after care is becoming increasingly standardised. It is currently a widely held standard that patients receive physiotherapy on the first post-operative day but some studies have already shown that aggressive physiotherapy and mobilisation on the day of surgery has reduced LOS following a TKR [13, 14]. There is also evidence to suggest that there is no change in the rate of complications following a TKR if mobilisation is initiated on the first post-operative day [15], thus highlighting the benefit of early and aggressive post-operative mobilisation.
Pre-operative haemoglobin and the first post-operative haemoglobin were both found to be negatively correlated with LOS. Interestingly, there was no association between the haemoglobin drop and LOS, suggesting that it is not the absolute blood loss, but rather the patients’ starting point that influences LOS. Since significant blood loss in joint replacement surgery can be anticipated and is a recognised complication, many measures are already in place as standard in order to address this, including application of a tourniquet and peri-operative blood salvage and autologous blood reinfusion. Since these results show that it is indeed the patients starting haemoglobin that is associated with LOS, this would suggest an added role for boosting patients’ haemoglobin pre-operatively, especially in anaemic patients, in addition to employing these blood conservation techniques. Studies have shown a role for pre-operative erythropoietin and iron supplementation in mildly anaemic patients undergoing surgery in which significant blood loss may be anticipated [16]. Other techniques that have been investigated in order to minimise blood loss include the use of intra-operative tranexamic acid and intra-articular epinephrine [17]. There is some evidence to indicate that it is possible to predict the need for blood transfusion following a TKR [18, 19], which would allow the above supplementary techniques to be employed in those patients, deemed to be at higher risk of requiring transfusion.
There are also some important negative findings. In our study, there was no relation between weekday of surgery and length of stay. This finding may be explained by the fact that physiotherapy is available Monday to Saturday at our centre. By contrast, other groups have shown that there is a difference and have attributed this largely to the lack of physiotherapy over the weekends [5]. We also found that there was no difference in length of stay between those patients operated upon in the morning compared to the afternoon. This finding is supported by other groups [5] and can be explained by the fact that all patients tend to be seen by the physiotherapists on the first post-operative day regardless of whether they have been operated upon in the morning or afternoon.
We also found that there was no difference in LOS between patients receiving a single shot femoral or combined femoral-sciatic nerve block and those who did not. This finding has been supported by many other studies [20, 21]. However, there are studies that have demonstrated that those receiving a single-shot femoral nerve block are able to ambulate earlier and are discharged earlier than those receiving a placebo [22].
Most studies have demonstrated that women tend to stay in hospital longer than men following a TKR in patients of all ages [5, 7, 23]. However, we found that in patients aged over 75 there was no significant difference in LOS between the male and female patients. This finding has been supported by a few other studies [24, 25]. Interestingly, we also found that in patients over 75, increased comorbidity did not correlate with increased length of stay. It is noteworthy however, that in patients of all ages, we found that comorbidity and ASA grade are both positively correlated with LOS. Other groups have replicated this finding [5].
While the finding that previous arthroplasty is not associated with increased LOS has been supported by some [24], others have demonstrated an inverse relationship between previous arthroplasty and LOS, postulating that by virtue of their previous experience, patients may gain greater benefit from the rehabilitation programme [7].
Our study does have a few limitations. First and foremost, since it was a retrospective study, there was a great dependence on the accuracy and the completeness of documentation in the medical notes. Second, since we have investigated such a large number of factors, there is an increased risk of a type 1 statistical error. However, even after Bonferroni correction for multiple statistical tests, the following factors were still found to be significant: pre-operative haemoglobin, post-operative haemoglobin, requirement for blood transfusion post-operatively, post-operative complications, post-operative day on which the patient was able to walk ten metres and achieve 90° active knee flexion. It is also possible that even though the discharge criteria used by each consultant are similar, there may be some subtle differences, resulting in slightly different discharge criteria depending on the consultant in charge.
In conclusion, it can be seen that the above factors are associated with post-operative length of stay in patients over the age of 75 undergoing TKR. Further investigation is warranted in order to determine the relative contributions of these factors to extended length of stay and propose a predictive model.
Acknowledgments
The authors would like to thank Professor Lee Shepstone of The University of East Anglia for his advice regarding statistical analysis.
References
- 1.Arden N, Nevitt MC. Osteoarthritis: epidemiology. Best Pract Res Clin Rheumatol. 2006;20:3–25. doi: 10.1016/j.berh.2005.09.007. [DOI] [PubMed] [Google Scholar]
- 2.Hospital Episodes Statistics Online (2009) Hospital episode statistics. Http://www.hesonline.nhs.uk. Accessed 14 February 2012
- 3.HM Treasury (2010) Budget 2010: Securing the recovery. Http://www.hm-treasury.gov.uk/budget2010.htm. Accessed 14 February 2012
- 4.Kaye KS, Anderson DJ, Sloane R, Chen LF, Choi Y, Link K, Sexton DJ, Schmader KE. The effect of surgical site infection on older operative patients. J Am Geriatr Soc. 2009;57:46–54. doi: 10.1111/j.1532-5415.2008.02053.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Husted H, Holm G, Jacobsen S. Predictors of length of stay and patient satisfaction after hip and knee replacement surgery: fast-track experience in 712 patients. Acta Orthop. 2008;79:168–173. doi: 10.1080/17453670710014941. [DOI] [PubMed] [Google Scholar]
- 6.Husted H, Hansen HC, Holm G, Bach-Dal C, Rud K, Andersen KL, Kehlet H. What determines length of stay after total hip and knee arthroplasty? A nationwide study in Denmark. Arch Orthop Trauma Surg. 2010;130:263–268. doi: 10.1007/s00402-009-0940-7. [DOI] [PubMed] [Google Scholar]
- 7.Dauty M, Smitt X, Menu P, Dubois C. Which factors affect the duration of inpatient rehabilitation after total knee arthroplasty in the absence of complications? Ann Phys Rehabil Med. 2009;52:234–245. doi: 10.1016/j.rehab.2009.01.001. [DOI] [PubMed] [Google Scholar]
- 8.Cook JR, Warren M, Ganley KJ, Prefontaine P, Wylie JW. A comprehensive joint replacement program for total knee arthroplasty: a descriptive study. BMC Musculoskelet Disord. 2008;9:154. doi: 10.1186/1471-2474-9-154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 10.Namba RS, Paxton L, Fithian DC, Stone ML. Obesity and perioperative morbidity in total hip and total knee arthroplasty patients. J Arthroplasty. 2005;20:46–50. doi: 10.1016/j.arth.2005.04.023. [DOI] [PubMed] [Google Scholar]
- 11.Hewitt B, Shakespeare D. Flexion vs. extension: a comparison of post-operative total knee arthroplasty mobilisation regimes. Knee. 2001;8:305–309. doi: 10.1016/S0968-0160(01)00110-7. [DOI] [PubMed] [Google Scholar]
- 12.Chandrasekaran S, Ariaretnam SK, Tsung J, Dickison D. Early mobilization after total knee replacement reduces the incidence of deep venous thrombosis. ANZ J Surg. 2009;79:526–529. doi: 10.1111/j.1445-2197.2009.04982.x. [DOI] [PubMed] [Google Scholar]
- 13.Isaac D, Falode T, Liu P, I’Anson H, Dillow K, Gill P. Accelerated rehabilitation after total knee replacement. Knee. 2005;12:346–350. doi: 10.1016/j.knee.2004.11.007. [DOI] [PubMed] [Google Scholar]
- 14.Larsen K, Sørensen OG, Hansen TB, Thomsen PB, Søballe K. Accelerated perioperative care and rehabilitation intervention for hip and knee replacement is effective: a randomized clinical trial involving 87 patients with 3 months of follow-up. Acta Orthop. 2008;79:149–159. doi: 10.1080/17453670710014923. [DOI] [PubMed] [Google Scholar]
- 15.Paysant J, Jardin C, Biau D, Coudeyre E, Revel M, Rannou F. What is the interest of early knee mobilization after total knee arthroplasty? Ann Readapt Med Phys. 2008;51:138–143. doi: 10.1016/j.annrmp.2007.11.003. [DOI] [PubMed] [Google Scholar]
- 16.Rosencher N, Ozier Y. Peri-operative use of EPO. Transfus Clin Biol. 2003;10:159–164. doi: 10.1016/S1246-7820(03)00057-0. [DOI] [PubMed] [Google Scholar]
- 17.Sasanuma H, Sekiya H, Takatoku K, Takada H, Sugimoto N, Hoshino Y. Efficient strategy for controlling postoperative hemorrhage in total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2011;19:921–925. doi: 10.1007/s00167-010-1263-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Salido JA, Marín LA, Gómez LA, Zorrilla P, Martínez C. Preoperative hemoglobin levels and the need for transfusion after prosthetic hip and knee surgery: analysis of predictive factors. J Bone Joint Surg Am. 2002;84-A:216–220. doi: 10.2106/00004623-200202000-00008. [DOI] [PubMed] [Google Scholar]
- 19.Guerin S, Collins C, Kapoor H, McClean I, Collins D. Blood transfusion requirement prediction in patients undergoing primary total hip and knee arthroplasty. Transfus Med. 2007;17:37–43. doi: 10.1111/j.1365-3148.2006.00698.x. [DOI] [PubMed] [Google Scholar]
- 20.Leach D, Bonfe M. The effectiveness of femoral/sciatic nerve blocks on postoperative pain management in total knee arthroplasty. Orthop Nurs. 2009;28:257–262. doi: 10.1097/NOR.0b013e3181b57aec. [DOI] [PubMed] [Google Scholar]
- 21.Duarte VM, Fallis WM, Slonowsky D, Kwarteng K, Yeung CKL. Effectiveness of femoral nerve blockade for pain control after total knee arthroplasty. J Perianesth Nurs. 2006;21:311–316. doi: 10.1016/j.jopan.2006.05.011. [DOI] [PubMed] [Google Scholar]
- 22.Toftdahl K, Nikolajsen L, Haraldsted V, Madsen F, Tønnesen EK, Søballe K. Comparison of peri- and intraarticular analgesia with femoral nerve block after total knee arthroplasty: a randomized clinical trial. Acta Orthop. 2007;78:172–179. doi: 10.1080/17453670710013645. [DOI] [PubMed] [Google Scholar]
- 23.Rissanen P, Aro S, Paavolainen P. Hospital- and patient-related characteristics determining length of hospital stay for hip and knee replacements. Int J Technol Assess Health Care. 1996;12:325–335. doi: 10.1017/S0266462300009661. [DOI] [PubMed] [Google Scholar]
- 24.Forrest G, Fuchs M, Gutierrez A, Girardy J. Factors affecting length of stay and need for rehabilitation after hip and knee arthroplasty. J Arthroplasty. 1998;13:186–190. doi: 10.1016/S0883-5403(98)90097-3. [DOI] [PubMed] [Google Scholar]
- 25.Liebergall M, Soskolne V, Mattan Y, Feder N, Segal D, Spira S, Schneidman G, Stern Z, Israeli A. Preadmission screening of patients scheduled for hip and knee replacement: impact on length of stay. Clin Perform Qual Health Care. 1999;7:17–22. [PubMed] [Google Scholar]