The efficacy of continuous-flow cryo and cyclic compression therapy after hip fracture surgery on postoperative pain: design of a prospective, open-label, parallel, multicenter, randomized controlled, clinical trial

The aim of the current study is to evaluate the efficacy of CFC therapy on pain in the first 72 postoperative hours of hip fracture patients. The secondary aim is to evaluate the effects on (non-) morphine analgesic use; postoperative hemoglobin level; transfusion incidence; delirium incidence and severity; use of psychotropic medication; length of stay; short-term location and duration of rehabilitation; functional outcome; short-term patient-reported health outcome; general and cryotherapy related complications. Furthermore the feasibility of a cryocompression device on orthopedic/surgery wards is assessed.

We hypothesized that: 1) CFC therapy will lower perceived pain levels and morphine consumption; and 2) will reduce postoperative blood loss and transfusion incidence; and 3) reduced pain by CFC therapy will lead to lower delirium incidence and 4) enhance functional recovery, leading to shorter length of stay in postoperative hip facture patients.

This study is designed as a prospective, open-label, parallel, multicenter, 1:1 randomized controlled, clinical superiority trial in accordance with CONSORT and SPIRIT guidelines [41, 42]. Eight orthopedic, surgery and/or geriatric departments of three middle sized teaching hospitals and one academic hospital in the Netherlands will participate: The Spaarne Gasthuis, Hoofddorp and Haarlem; Noordwest Ziekenhuisgroep, Alkmaar; Amstelland Hospital, Amstelveen; VU University Medical Center, Amsterdam.

A balanced 1:1 block (size six) randomization stratified by center will be performed directly after surgery, using Research Manager (RM; version 4.5.0.1), a web-based computer program (Fig. 1). A research nurse who is not involved in the study generates the allocation sequence with RM. Intramedullary hip nail (IMHN), dynamic hip screw (DHS) and cannulated screws are grouped for block randomization, as are total hip arthroplasty (THA) and hemiarthroplasty (HA). The intervention group ‘A’ will receive CFC therapy postoperative; the control group ‘B’ will receive standard care. Due to the nature of the study no blinding can exist as patients, family and physicians notice the use and settings of CFC therapy. Physicians randomize participants with a digital logon to the web-based system or the physician can contact the coordinating investigator and he will randomize the included participant (see Table 1 for criteria).The Informant Questionnaire for the Cognitive Decline in the Elderly (IQ-CODE) questionnaire is administered when the clinician has doubts about the cognitive state that may compromise CFC therapy sessions or study measurements (Table 1). Ideally a MMSE would be used for this purpose, however due frequently perceived pain at the accident and emergency A&E department and consequently administered narcotic analgesics, the hetero anamnestic IQ-CODE is used instead. The coordinating investigator contacts the participating hospitals on a daily basis by telephone or physically to ensure adequate and appropriate enrollment of participants and completeness of data. A list of excluded patients will be drafted.

On arrival at the accident and emergency (A&E) department patients receive acetaminophen intravenously, diclofenac (if not contraindicated) and morphine subcutaneously or fentanyl intravenously until numerical rating scale (NRS)-pain scores have dropped below 4. Upon admission the local protocol is started (Table 2). At all centers the 'as needed' medication is administered when the NRS-pain score is 4 or higher and no excessive (opioid-induced) sedation is present. Patients who are deemed able are given intravenous patient controlled analgesia (PCA) pumps with morphine (Table 2). In patients older than 70 years the bolus setting is reduced by 50 %. Preferably patients are operated on using spinal anesthesia. In order to put the patient in an upright position for spinal anesthesia administration femoral blocking with short-acting analgesics can be used. All centers use bupivacaine 0.5 % between 2.0 and 3.0 ml administered at the lumbar level. Additional (long acting) analgesics administered during surgery are noted. The participating centers adhere to the Dutch national guidelines for surgical technique for the various fracture types [43]. Postoperatively, on the day of surgery or the first postoperative day physical therapy is commenced once or twice daily. Physical therapy sessions are usually 30 min long. In the period in which the focus of this study lies physical therapy focuses on strengthening of quadriceps and gluteal musculature, walking and making transfers.

Continuous-flow cryocompression therapy is applied by using the ‘Game Ready system’ (GRS; CoolSystems Alameda, California). Through an anatomically designed hip/groin wrap covering most of the thigh and pelvis up to the iliac crest, the GRS simultaneously delivers both adjustable continuous-flow cryotherapy and intermittent compression through a portable control unit filled with ice and water. The machine has four pressure settings: no pressure, low pressure (5–15 mm Hg), medium pressure (5–50 mm Hg) and high pressure (5–75 mm Hg). Temperature can be adjusted between 4 and 13 °C and is indicated by one, two or three snowflakes. The lowest temperature is started and maintained throughout the study if feasible. Pressure is started at ‘low’ and is increased one step per 4 treatments (Table 3). Depending on the end of surgery patients will be categorized in three treatment schedules with respective start and end-times (Table 3). If patients are uncomfortable the appropriate adjunct will be adjusted stepwise until a comfortable setting is reached, deviations are noted. Adjustments are recorded and comfortable settings are maintained according to the discomfort flowchart (Fig. 2). Patients will be treated between 10 and 12 times during the first 72 postoperative hours, each cycle lasting 30 min. Preferably, treatment cycles and control measurements during the first 72 postoperative hours are performed at fixed moments: 8:00 h, 12:00 h, 16:30 h, and 21:30 h. The GRS-wrap is only in place when CFC therapy is administered and applied/removed by the nurse.

In case of discomfort the appropriate adjunct is adjusted accordingly (Fig. 2) and supplemental analgesics are given as needed (Table 2). In both the intervention group receiving CFC therapy and in the control group no restrictions are made towards regularly used static compressive bandages and/or wound drains (with or without autologous re-infusion). Geriatricians are consulted in a standard fashion in patients aged 70 years and over.

Patient demographic data are noted on admission as well as American Society of Anesthesiologists (ASA) class, current NRS-pain, delirium observation screening (DOS)-score, activities of daily living (ADL)-Katz score and the New Mobility Score (NMS; Table 4) [44‐46]. Time of injury and arrival at the A&E department is noted. Patient delay intervals longer than 6 h are actively explored and reasons for delay are clarified e.g. if a patient was prone and unable to call emergency services.

A standard preoperative venous blood sample is taken in which hemoglobin, hematocrit and mean corpuscular volumes values are measured (Table 4). Upon admission the local hospital analgesic and antithrombotic protocol is started (Table 2). A geriatrician is consulted preoperatively in patients who are older than 70 years and will focus on: polypharmacy, risk reduction of falls, prevention and/or treatment of delirium, optimization of nutritional status. The ADL-Katz score and delirium risk assessment score (DRAS) [47] are completed upon admission; furthermore precipitating factors for delirium (bladder infection and/or retention, fixation, sleep deprivation, electrolyte abnormalities including renal function impairment) are noted throughout the hospital stay. After the operation the surgeon notes the intraoperative blood loss, the type of implant, if applicable intraoperative complications, randomizes the patient and determines the patient’s treatment schedule.

Intervention patients start with CFC therapy 6 h postoperative, until 72 h postoperatively. A trained ward nurse will conduct the pain assessments when the patient is in bed or stationary in a chair for at least five minutes. Directly before and directly after CFC therapy vital signs including NRS-pain are measured by the ward nurse, when the patient is prone in bed for at least five minutes before commencing and five minutes after cessation of CFC therapy. In control patients only NRS-pain is measured at the same time as intervention patients would normally be assessed. All pain measurements are continued throughout the first postoperative 72 h for control and intervention patients alike (Table 4). Three times daily DOS-scores [48] will be obtained; in case of a DOS-score higher or equal to three, geriatricians diagnose and subsequently monitor delirium severity with the delirium rating scale revised 1998 (DRS-R-98) [49] on a daily basis or until the score drops below 12.25 out of 39 (Table 4). Time and amount of administered (psychotropic) medication will be documented. The nurse and ward doctor inspect the wound and dressing in accordance with routine care. Blood samples are taken at postoperative day one and three, and post-transfusion if applicable (Table 4). The timed up and go test (TUGT) is performed before discharge (Table 4). The TUGT will be administered in all patients who are able. If no weight bearing is allowed or patients are physically unable, than the test will be postponed to the outpatient visit. The mini mental state examination (MMSE) is administered for data stratification purposes [50‐52], if patients are delirious during admission the MMSE is postponed to the outpatient visit.

In order to document experiences with CFC therapy all treatment patients will fill in a questionnaire at discharge, which is specifically drafted for this study (Table 5). During the study a booklet will be made available into which staff can write down prevalent occurring technical GRS-related problems.

At the single outpatient study visit between six to eight weeks, the following parameters are assessed; current NRS-pain; analgesic use; mobility status by the TUGT, de Morton Mobility Index (DEMMI) and the Short Physical Performance Battery (SPPB); current living situation (rehabilitation or at home); wound status and complications (thrombosis, consolidation, infection, readmission). Furthermore the self-assessment health questionnaires EuroQol(EQ)-5D-3 L and Short Form (SF)-12 are completed. If not administered during admission the MMSE is completed (Table 4).

Upon inclusion patient characteristics are entered in RM and postoperative randomization takes place. A unique patient number is generated by RM, which will be used for further data collection on preformatted case report forms (CRF’s). Only the coordinating investigator has access to the key-file that can identify patients from RM-generated patient numbers, it will be stored at the coordinating center throughout and after completion of the trial and will not be shared. The Spaarne Gasthuis is the unrestricted owner of the final dataset; no contractual agreements exist in regard to publication policies. No interim analysis will be performed.

The coordinating center comprises of: the principal investigator, the coordinating investigator, a research nurse, a research assistant, a clinical epidemiologist and statistician. The research assistant enters data in RM. The research nurse can be contacted via telephone to solve any (acute) problems that might occur and enters data in RM. The epidemiologist and statistician assist with data analyses and drafting of the manuscript. During the enrollment phase of the trial the coordinating investigator meets bimonthly with the individual principal investigators of the respective centers to assess the current inclusion rate, data quality, protocol adherence and prevalent occurring problems. During enrollment two meetings are planned to collectively assess problems and deliberate on how to instigate solutions to these problems. All participating personnel are updated on a monthly basis via a newsletter in which current inclusion rates are presented, and to point out points of attention and solutions to frequently observed problems. During physical visits CRF’s are collected, assessed for completeness and missing data is retrieved from patient medical records where possible; secondly the appropriateness of administered treatments and (functional) tests is assessed.

Two independent data monitors are instigated and will monitor the trial at two of the five sites where highest inclusion rates are expected. Results of the monitor will be discussed in a committee independently of the investigators and the investigators will be informed both orally and in writing. Since the overall risk of participation in the trial is considered low by the REC/IRB no external safety overseers are required by national guidelines. If selected, external auditing is possible.

The verbally administered 11-point (range 0–10) NRS-pain is a widely accepted assessment tool and has been validated in many different conditions, among is acute pain in the emergency department [53‐55]. Reliability and validity has been thoroughly studied in (cognitively impaired) elderly and has shown to maintain acceptable clinometric characteristics [56‐58]. NRS-pain scores are compared within the intervention group i.e. pre and post-treatment and compared between the intervention and control groups. Cumulated pain scores will be calculated for both groups by adding the (post-treatment) pain scores of the first 72 postoperative hours.

All postoperative oral and parenteral administered narcotic analgesics will be converted and added up using accepted algorithms [59, 60] and compared after the last treatment day, at discharge and at the outpatient clinic. Total milligrams of acetaminophen and non-steroidal anti-inflammatory drugs (NSAID’s) are reported in the same fashion. At the outpatient visit patients will be asked about their current pain medication use.

The hemoglobin values from the obtained blood samples at postoperative day one and three will separately be detracted from the preoperative concentration and postoperative day three will be detracted from postoperative day one. An adjustment will be made for the intraoperative blood loss. Indications for transfusion are managed according to the Dutch national blood transfusion guideline [61]. Normovolemic ASA-one patients older than 60 years who lose blood at a single locus will be transfused at 5 mmol/l (8 g/dl). Patients unable to elevate cardiac output for hemodilution are transfused at 6 mmol/l (9.6 g/dl). Incidence of erythrocyte transfusion is noted and compared between groups.

The DRS-R-98 is considered to be a valid and reliable instrument for delirium diagnosis and documenting delirium severity [49, 62]. It takes a trained nurse or physician five to ten minutes to assess the patient and comprise the DRS-R-98 score, and is ideal for longitudinal delirium follow-up. The sensitivity and specificity of the scale are reported to be 92 % and 86 % respectively. DRS-R-98 scores are compared within and between groups, and newly started psychotropic medication is compared.

Admission time is measured and expressed in two ways: from arrival time at the A&E department until discharge and from end of surgery to until discharge. The discharge criteria in all participating centers are: NRS-pain score below 5 without need for parenteral analgesics and an Elderly Mobility Scale (EMS) score 14 out of 20 or higher [63, 64]. If this some items were deemed unobtainable EMS scores are completed until transfer to a nursing home is arranged. If family members are able to foresee in certain home care aspects the patient can be discharged to his/her home. The fulfillment of this care by family is noted and corrected for afterwards.

The surgeon monitors all complications that may occur throughout admission, in general or cryotherapy-related. After discharge the coordinating investigator verifies if all adverse events are registered and recorded on the appropriate CRF. In case of serious adverse events relating to the treatment patients will no longer receive treatment but will remain in the study for follow-up, if feasible. Severe leakage i.e. multiple bandage swaps daily is noted and registered as an adverse event. Furthermore an intragroup analysis of vital signs is made to determine possible central effects of the administered therapy. Finally, at the outpatient visit patients are assessed for adverse events that may have occurred during rehabilitation. Additional to the patient questionnaire (Table 5), intervention patients are asked about their experiences with CFC therapy. An additional insurance is taken out to financially compensate participants if applicable. Adverse events are tabulated and reported in the final manuscript.

At the end of the study nursing staff will complete questionnaires specifically drafted for this study (Table 6) and the booklet of prevalent technical problems will be evaluated. A close out visit will be planned in which recommendations for further improvement of the machine will be discussed. Treatment failure i.e. discontinuation of treatment is reported and provided with reasons e.g. discomfort.

All statistical analyses will be computed using the SPSS statistical package (IMB SPSS, Inc., Release 20.0.0.0, 64-bit edition). Statistical analysis will be performed according to the intention-to-treat principle. Baseline characteristics will be described in accordance with CONSORT guidelines [41] using means and standard deviation in case of normal distribution, and medians and interquartile ranges otherwise. Continuous variables will be checked for normality, visually and by using the Shapiro Wilk test. Our primary analysis focuses on the differences in NRS-pain scores between the study groups 24 h postoperative and will be analyzed by use of Student’s T or Mann–Whitney U tests in case of skewed distribution. To assess treatment effect over time, a mixed model analysis for repeated measures will be performed. This model allows missing data and adjustment for serious confounders and interactions. In case of skewed distributions and outliers non-parametric variants will be used (e.g. Mann Whitney U or Friedman tests). Secondary continuous outcome measures (EQ-5D-3L, SF-12, DRS-R-98 and hemoglobin) will be compared by use of Student’s t or Mann Whitney U tests. Ordinal variables will be analyzed using the Mann–Whitney test. Chi-squared tests will be performed in case of categorical variables. In case of important confounding, analysis will be adjusted to correct for these factors (by use of multiple regression analysis, linear or logistic). A p-value of < 0.05 will be considered statistically significant.

Sample size calculation is based on the primary outcome measure, NRS-pain at 24 h postoperatively. Subsequently, a mean NRS-pain score for postoperative day one was drawn from the Spaarne Gasthuis electronic patient dossier digital databanks and a standard deviation of 2.2 was calculated for pooled internal fixation and (hemi) arthroplasty for cervical and peritrochanteric fractures. The minimal clinical relevant difference for NRS-pain is 1.3 out of 10 [53]. The alpha was set at 5 % with a power of 90 % due to anticipated heterogeneity in regard to operating techniques, general care protocols but predominantly in regard to pain protocols between the participating centers (Table 2). An expected NRS failure rate of 10 % is anticipated together with expected missing data a drop out of 22.3 % is computed. The total number of included patients will be 160.

Permission has been obtained from the 'Medisch Ethische Toetsingscommissie Noord-Holland' regional medical ethics committee (NL45657.094.14) and the trial is registered at www.​trialregister.​nl on the 23^rd of August 2013 with trial number: “NTR4152”. Patients with an intra or extracapsular hip fracture who meet the inclusion criteria and without apparent exclusion criteria (Table 1) are informed about the study both orally and in writing by the assessing physician e.g. the local ward doctor, staff member or consulting geriatricians. Depending on operating room capacity and patient characteristics patients will either first be transferred to the ward or will go directly to the operating room. Patients can be included pre and postoperatively within 6 h post surgery, before the first treatment starts 6 h postoperative (Fig. 1). Maximum time for consideration to participate in the study is set at 12 h. Patients approving to participate need to sign the informed consent form after which baseline characteristics are noted.

Not applicable.

Not applicable as this is a protocol of a study and no data are reported.