Prognostic Value of Baseline High-Sensitivity C–Reactive Protein in Patients Undergoing Replacement Arthroplasty

Introduction: A prospective hospital based blinded study was carried out in an orthopedics department in a medical college hospital of Kolkata, West Bengal, India, with the objective of assessing the role of pre-operative high-sensitivity C-reactive protein (hsCRP) level in predicting postoperative complications in replacement arthroplasty. Methods: One hundred and twenty one study subjects were selected. The blood collected in pre-operative, perioperative and postoperative states were stored and ﬁ nally analyzed after the data collection was over after the stipulated follow-up time of post operative Day-14. The patients were divided into two groups according to baseline hsCRP level <3mg/dl and >3mg/dl. The other preoperative parameters measured were found to be statistically comparable. Results : Among important ﬁ ndings, operative time was signiﬁ cantly higher in the group with hsCRP>3mg/dl, postoperative complication rate was also signiﬁ cantly more in that group. Postoperative complication in Day-7 though quite high among both the groups (45% and 50.9%), the difference was not statistically signiﬁ cant. Comparison of hsCRP levels at baseline, perioperative, postoperative D7 and D14 also showed no signiﬁ cant changes. Conclusions: We concluded that high hsCRP level(>3mg/dl), in patients with apparently no other risk factors may be at higher risk of developing complications after post operative D14 of replacement arthroplasty and operative time also signiﬁ cantly increases with high baseline hsCRP level.


INTRODUCTION
High sensitivity C reactive protein (hsCRP) has been considered as an important predictor of cardiovascular risk.It's level is consistent across gender and ethnic groups in the population. 1 Many studies shows that increased CRP level signifi cantly predicts occurrence of future myocardial infarction, stroke, cardiovascular death and peripheral arterial disease.There is a linear 'doseresponse' association between CRP and subsequent vascular events, which are independent of the traditional risk factors (age, smoking, hypertension, diabetes and dyslipidaemia) evaluated in daily practice. 2A three fold higher incidence of cardiovascular disease is found in persons having CRP>3mg/dl when compared to those with <3mg/dl.CRP levels are also inversely proportional to VO2max, an important marker of cardio respiratory fi tness. 3,4Impaired cardio respiratory reserve is strongly associated with poor peri-operative outcome. 5The present study aims to fi nd out whether higher levels of pre operative hsCRP is associated with increased perioperative morbidity and mortality in patients undergoing elective, intermediate risk replacement arthroplasty.

METHODS
A prospective cross-sectional study was conducted in which pre and post operative (baseline, Day-7 and Day-14) serum hsCRP level has been estimated in all the subjects in the departments of Orthopedics and Anesthesiology of Calcutta National Medical College and Hospitals in collaboration with departments of Pharmacology and Community Medicine, Medical College, Kolkata West Bengal, India from December, 2007-May 2008.Due permission of the Institutional Ethical Committee of Calcutta National Medical College was obtained.Study subjects were enrolled from the patients presenting for elective joint replacement surgery including revision procedures (Total Hip Replacement, Total Knee Replacement and hemi-arthroplasty) in both outpatients' department and inpatients' wards of the Department of Orthopaedics, Calcutta National Medical College after they fulfi lled the inclusion and exclusion criteria laid down for the study.Written informed consent was duly obtained from all the participants in their own language.
The patient with either sex, requiring elective joint replacement surgery, revision procedures (Total Hip Replacement, Total Knee Replacement and hemiarthroplasty), fi t for major elective surgery as opined by orthopaedician and anesthesiologist and Willingness to give written informed consent for the study procedures were included in the study.The patients with high cardiac risk factors (BP>160/110 mm of Hg; Grade III, IV angina; uncontrolled heart failure etc.), signifi cant renal impairment (creatinine clearance<30 ml/min), signifi cant hepatic impairment (ALT>200μ g/dl; Prothrombin Time>15 seconds), clinically signifi cant physical and mental abnormalities were excluded.
All the patients satisfying the selection criteria were recruited for the study during these six months and after fi ve patients were excluded due to exclusion criteria, fi nally 121 patients became the study subjects.
Brief history of the subjects was taken and relevant general examinations; laboratory reports of haemoglobin, urine for albumin were recorded.Peripheral venous blood samples were drawn from all patients via a forearm vein in the pre-admission clinic, a mean (SD) of 12 (7) days prior to surgery.Samples were processed immediately, coded and stored at -800 C until blinded, independent analysis after all clinical data collection had fi nished.Serum hsCRP level was measured using immunoturbimetric method.The collected blood samples were fi nally analyzed at the end of study to avoid bias.Preoperative, perioperative and two post operative (day 7 and 14) were collected from the patients.
After analysis of hsCRP level, patients were divided into hsCRP groups low (<3mg/dl) and high (>3mg/ dl) according to the Centre for Disease Control Classifi cation. 6rgery was done under general or regional (epidural or spinal) anesthesia.Assessment of total blood loss done by measuring peri-operative and post-operative drain collection and blood transfusion was done accordingly to compensate the loss.Peri-operative cardiovascular risk index was calculated according to American College of Cardiology guideline 7  Evidence based peri-operative factors associated with peri-operative morbidity like body mass index, 8 operative time, 9 post-operative haematocrit, 10 immediate post operative temperature, 11 were also noted prospectively.Post-operative morbidity (POM) was assessed using a validated system. 12Various parameters were tested among the two groups for statistically signifi cant differences.Unpaired 't' test and chi square tests were used as indicated and P<0.05 was considered to be signifi cant.

RESULTS
Preoperative characteristics of subjects grouped according to hsCRP level < or >3mg/dl (Table 1).Comparison of preoperative medication among the study subjects grouped according to hsCRP level < or >3mg /dl was done in the cardiac medication status (Table 2).Out of the total 121 patients 60 preoperative samples showed baseline hsCRP level <3mg/dl (mean-1.76)and rest 61 had hsCRP level >3mg/dl (mean-10.5).They were grouped in two different categories and further analysis was done based on this classifi cation.Henceforth, the group with hsCRP <3mg/dl will be referred to as fi rst group and the group with hsCRP >3mg/dl will be referred to as second group.Both the groups showed some similarity in characteristics i.e. male predominance (60% and 62.3%), comparable mean BMI (22.3 and 25.3, P>0.05), a good proportion of smokers (45% and 54.1%) and hypertensive (51.7% and 59.02%).Other co-morbid conditions like diabetes mellitus (11.7% and 14.8%) and chronic obstructive airway diseases (10% and 14.8%) were also comparable in number.

Ghosh et al. Prognostic Value of Baseline
High-Sensitivity C-Reactive Protein in Patients ... Intake of cardiac medications was also noted and it showed that majority of the patients in both the groups were already under therapy.First group with hsCRP<3mg /dl had 53(88.3%)and the other group had 51(83.6%)patients with cardiac medication.The difference was not statistically signifi cant.Aspirin was the commonest drug in both the groups (37.7% and 54.1%), followed by ACE inhibitors (21.7%and 32.8%).In fi rst group next commonly used drug was steroids (20%) while in the other it was beta blockers (29.5%).However when we analyzed the medication use pattern in two groups the difference was not statistically signifi cant.
In a group with hsCRP<3mg /dl, equal number of patients underwent operation with general and regional anesthesia.In the other group majority had regional anaesthesia (62.3%) (Table 3).Operative time was signifi cantly higher in case of patients with hsCRP>3mg /dl (107.3min.compared to 88.5min in other group, P<0.05 value-0.0001).Other parameters measured though dissimilar, never were found to be statistically different.
At post-operative day 14 (D14), number of patients with complication, were greatly reduced (2/60-3.3% in fi rst group and 10/61-16.4% in second group).The difference was statistically signifi cant.Infection was the commonest (3.3% and 11.5%) complication, followed by GI disorders (3.3% and 4.9%) and cardiac complication (1.7% and 3.3%) other disorders noted were neurological problem (one in the fi rst group) and renal impairment (one in the second group)

DISCUSSION
HsCRP has established itself as a strong predictor of cardiovascular pathology in general non surgical population.Our results from a group of surgical patients at risk of peri operative complications and cardiovascular dysfunction are consistent with these reports.Similar results have been published which highlights the potential importance of preoperative hsCRP in predicting outcome in esophageal resection 13 and vascular surgery. 14The present study aims to fi nd out whether higher levels of pre operative hsCRP is associated with increased peri and post operative morbidity in patients undergoing elective, intermediate risk replacement arthroplasty.The results reveal that hsCRP is defi nitely an important predictor for post operative complications at D14, but evidence is not enough for its relevance in predicting earlier complications.Some investigators speculated that very high hsCRP >10 mg/dl could be due to acute phase response and might give false positive results, later data showed this to be unlikely. 1Orthopaedic patients with chronic, high grade infl ammation having persistent high hsCRP >10 mg /dl are at higher risk.Further, the preoperative CRP levels in patients undergoing revision procedures in our study were not disproportionately distributed in the higher CRP group which matches results of previous reports. 15

Figure 1. CRP levels
Perioperative outcome is strongly associated with cardio respiratory fi tness which can be determined by cardiovascular exercise testing and myocardial ischaemia arising out of it which can predict complications in higher risk surgical patients. 5Various studies substantiate the transient increased risk of myocardial infarction after infection. 16Again, chronic infl ammatory conditions and autoimmune diseases like systemic lupus erythematosus and rheumatoid arthritis are at increased risk of cardiovascular complications. 17There is steady association of hsCRP and cardiovascular complications in different patient groups, 3,4 so estimation of hsCRP may be a useful screening tool to foresee perioperative risks, period of hospitalization and potential costs.
Line diagram comparing CRP levels show that there is no signifi cant change in baseline, seventh and fourteenth postoperative day.The samples were collected within 2-3 weeks of operation, a time span over which majority of reports suggest hsCRP is unlikely to change. 18However regardless of fl uctuations in hsCRP levels, several studies emphasize that intermittent, chronic and acute infl ammation may all be associated with cardiovascular dysfunction.The main pillar of the study is the fact that the patients, investigators and healthcare personnel were blinded to the key data i.e. hsCRP.So, it is unlikely that clinical assessment by the investigators were anyway biased to diagnose post operative morbidity in the study subjects.As peri/post operative morbidity is always an unacceptable outcome for surgical procedures, measuring and manipulating certain parameters pre and peri operatively may reduce/predict the adverse events and fi nally benefi t the quality of patient care in the hospital.
There is lack of signifi cant differences in post operative complications in both the groups on the seventh postoperative day.This may be due to the fact that only qualitative refl ection of various complications has been made and we have not recorded the degree of severity of these complications.All of the complications are not directly related to cardiovascular dysfunction (and hence hsCRP) but the commonest complication is gastro-intestinal side effects for which peri-operative gut hypo perfusion is at least partially responsible. 19n contrast, both the incidence of complications 10/12 (P<0.05) and prolonged hospital stay was observed in the higher hsCRP group on the fourteenth day.

CONCLUSIONS
Baseline hsCRP demographics may prove to be a useful tool in different surgical populations because the underlying pathology requiring surgery may be an important determinant of preoperative hsCRP levels.We can assume that surgical complications may not be due to the magnitude and complexity of surgical interventions alone but also monitored by underlying chronic infl ammatory process.The data obtained by this study suggests that hsCRP merits further study in the perioperative setting for specifi c surgical groups.
Preoperative measurement of hsCRP may be a fruitful strategy to stratify perioperative care.
by weighting low (x1) and intermediate factors (x2): No patients had high risk factors.
Ghosh et al.Prognostic Value of Baseline High-Sensitivity C-Reactive Protein in Patients ...

Table 5 . Comparison of post operative morbidity on D-14 among study subjects grouped according to hsCRP level
Ghosh et al.Prognostic Value of Baseline High-Sensitivity C-Reactive Protein in Patients ...