Elderly dehydrated patients (age\>70 years) with proximal femur fracture scheduled for orthopedic surgical repair under spinal anesthesia were recruited. Patients' medical history, physical examination, electrocardiogram (ECG), and X-ray evaluation were performed during the preoperative assessment and supplemented by specific exams (e.g., troponin, pro-BNP levels, coronary angiography, myocardial perfusion scintigraphy, echocardiographic examination, or stress tests) as per consultant cardiologists' recommendations. Informed consent was obtained from the patients or their surrogates.
Based on cardiology consultation, the initial patient screening was used to identify those patients who were American Heart Association/American College of Cardiology (AHA/ACC) stage B or C whose cardiac disease status was compensated. Patients with tachycardia (heart rate \> 100 beats/min), atrial fibrillation or left bundle branch block on ECG and those who diagnosed with moderate to severe or severe tricuspid or mitral valve regurgitation, severe aortic or mitral stenosis, severe pulmonary hypertension, right heart failure were not recruited. Patients who fulfilled the above criteria were scheduled for TTE examination prior to spinal anesthesia.
Baseline blood pressure measurements. In all the participants, the cardiovascular medication except for beta blockers was not given the day of surgery. Medications with standardized dose time once a day (e.g. angiotensin converting enzyme inhibitor) were discontinued 24 hours before surgery. Upon arrival in the operating room (preparation area), standard non-invasive monitoring (continuous ECG, non-invasive blood pressure measurements every three minutes and SPO2) was applied. Prior to placement of 20/18G intravenous and indwelling radial artery catheters, three consecutive MAP measurements were recorded, and their average was used as reference value.
Preoperative echocardiographic examination. After baseline MAP measurements and prior to spinal anesthesia induction, a complete TTE examination was performed which included the following views: parasternal long (LAX) and short axis (SAX), apical 4-chamber (4CH), including also the "RV-focused view", apical 2-chamber (2CH), apical 3-chamber (3CH) and subcostal IVC (SUB-IVC). All data were saved for off-line analysis. The physicians who analyzed the data were all competent at TTE. The LV-EF was estimated by the Simpson's method of discs by performing measurements of LV volumes in the 2CH and 4CH views. In suboptimal images, ultrasound contrast agents were used for the improvement of endocardial border definition. (SonoVue: Bracco, Milan, Italy and Optison: GE, Helthcare, USA).
To assess the function of the right ventricle (RV), the "RV-focused view" and the SUB-IVC views were acquired. The FAC, (2-dimentional surrogate for RV-EF) was calculated by the following formula: 100 × (RV-Area end-diastolic - RV-Area end-systolic) / RV-Area end-diastolic). Tricuspid regurgitation grading and velocity were assessed according to pertinent guidelines. LV filling pressures were estimated by the E/Em ratio (E=peak velocity of mitral flow in early diastole, Em=the average of peak velocities in early diastole of lateral and septal mitral annulus). Stroke volume (SV) and subsequent stroke volume index (SVI=SV/m2) of the LV were assessed using automated measurements of LV volumes, according to the formula SV = EDV-ESV, where EDV=end diastolic LV volume, and ESV=end systolic LV volume. From these data, we derived values for the assessment of cardiac output (CO) =SV x HR, and SVR =MAP x 80/CO, (MAP=mean arterial blood pressure, HR=heart rate). The IVC measurements included the IVC maximum diameter at the end of expiration (dIVCmax), the IVCCI during spontaneous, quiet, breathing \[(IVC maximal diameter - IVC minimal diameter)/IVC maximal diameter\], and the ratio of dIVCmax-to-IVCCI. All data were saved and stored digitally for off-line, postoperative analysis. All measurements were repeated twice (by a physician who were unaware of patients' hemodynamic status), and their arithmetical mean was taken for the analysis.
Spinal anesthesia protocol. As per departmental fasting guidelines, all patients were allowed to receive clear liquids and solid food until 2 and 8 hours before induction to spinal anesthesia, respectively. No pre-spinal anesthesia fluid load was given.
Ultrasound-guided fascia iliaca or supra-inguinal fascia iliaca nerve blocks were performed before spinal anesthesia to prevent patients' pain during lateral position placement. Ten to 20 mg of propofol were administered prior to nerve blocks. Additional sedation/analgesics (e.g propofol or remifentanil infusion) were not administered during the operation.
Spinal anesthesia was introduced with a single intrathecal injection of ropivacaine (0.75% solution) ranging between 12 to 18 mg (average dose 15 mg) depending on patients' age, frailty, and BMI, using a 22- or 25-gauge needle. The intrathecal injection was performed in inter-spaces L2-L3, L3-L4 or L4-L5. After subarachnoid injections, all patients immediately returned to supine position except for those with hip hemi-arthroplasty. Spinal anesthesia was followed by a continuous intravenous infusion of 1.5 ml/kg/h of Lactate Ringer's solution. Sensory block was assessed every 10 minutes after injection (pinprick test) until the end of surgery. The maximum cephalad dermatomal extent of the sensory block was measured. Spinal anesthesia was followed by a bolus 8 mgr/kg and a continuous intravenous infusion of 1.5 ml/kg/h of Lactate Ringer's solution.
Intraoperative blood pressure measurements and patients' follow up. After spinal anesthesia induction, continuous, invasive blood pressure measurements were taken and patients with MAP ≤ 65 mmHg, or with a reduction ≥ 25% of baseline pre-operative values, were considered hypotensive (time interval of low MAP: 30 seconds, assessment period of low MAP: from spinal anesthesia induction to the end of surgery). Once the patients experienced hypotension (fulfilling the above criteria) the patients were treated according to the attending's anesthesiologist clinical judgment.
The length of stay (LOS) in the PACU was also estimated and defined as the time from the patients' admission to the time of their discharge from the PACU. Patients who were hemodynamically stable during the intraoperative period, but experienced hypotension in the PACU, were not considered in the statistical analysis.
