Hip fracture patients are considered a frail group of patients and are seen on a daily basis in the Emergency Department (ED). Hip fractures occur at around 1.6 million per year worldwide, with this number expected to rise to 4.5 million by 2050 because of progressive aging [1]. It concerns an injury in need for operative treatment, due to the known and potential lethal complications. The 30-day mortality is high and is found to range between 7.5–13.3% [2,3]. The cause of death is rarely due to the primary surgical condition, but as a result of medical complications during or after admission. The most common causes of 30-day mortality are pneumonia, sepsis and myocardial infarction [2].

Because of the frailty and the high postoperative mortality in this population the preoperative management is vital in order to identify correctable comorbidities and optimize the clinical condition of the patients as much as possible. The preoperative screening exists of a thorough history taking, physical examination, additional blood sampling, an electrocardiogram (ECG) and further additional testing if indicated. However, correcting comorbidities does cost time with possible worsening of patient’s condition due to immobilization. The timing of surgery remains vital but still some controversy exists. Unnecessary diagnostics with no consequences for the perioperative care that could delay surgery are potentially harmful and should be kept to a minimum.

Preoperative chest radiographs (POCR) are used in many hospitals in the preoperative setting, especially to rule out preexisting chest infections and evaluate existing disease to reduce the risk associated with anesthesia and surgery, increasing preoperative care quality [4]. However, the additional value of the POCR is under debate. An abnormal POCR can influence anesthetic and surgical management and can delay or adjourn the operation. The limited research on the POCR so far has shown that few POCR abnormalities demand a change in anesthetic or surgical management or correlate with a negative outcome during hospital stay in general surgical conditions [5]. Several studies [4, 5, 6, 7, 8, 9] and more recent international guidelines [10, 11] suggest that the additional value of POCR is limited in patients undergoing elective non-cardiothoracic surgery and should only be performed when clinically indicated, in immigrants from developing countries without a chest radiograph in the last 12 months [4, 5,6, 7, 8, 9, 10] or in patients scheduled for critical care [11]. Despite current recommendation the routine POCR is still frequently performed [4]. Because of the limited quality of evidence no clear recommendations can yet be made, and research is needed to identify the best strategy for specific surgical conditions in specific patient populations [4]. No studies have yet described the value of POCR with regard to the perioperative care in nonelective surgery patients or hip fracture patients specifically. Most of the hip fracture patients undergo POCR in the VU University medical center per protocol. With the rising age and rising incidence of hip fracture patients [1, 12], with a POCR costing €50, a significant cost reduction could be made by performing selective POCR without compromising the quality of perioperative care. This study assesses the additional value of routine POCR in hip fracture patients and will address the findings on the POCR, their influence on the perioperative management and the relation between per- and postoperative complications.

A retrospective analysis of a database of patients aged 18 years and older sustaining low-energy trauma (LET) treated by a trauma surgeon for a hip fracture in the VU University Medical Center (level 1 trauma center) within a five-year period (2008–2013) was performed. Patients sustaining a high-energy trauma (HET) and patients aged under 18 years were excluded.

Patient’s characteristics regarding age, gender, mechanism of injury, type of hip fracture, American Society of Anesthesiologist (ASA) score, cardiac and/or pulmonary medical history, type of operation and all per- and postoperative complications during hospital stay were registered. All preoperative diagnostics and consultancies were reported and analyzed. If an operation was adjourned, the indication was assessed. All data were obtained by retrospective patient chart reviewing. The POCR results were classified as either normal or abnormal by the first author based on the report made by the radiologist at the time. If judged abnormal, the abnormality was registered. Elongation and calcification of the aorta were judged as normal. Preoperative screening results were checked, and patients were declared fit for surgery by an anesthesiologist. The collected information was processed in a SPSS^®20 database. Statistical differences were established by calculating the Pearson Chi-square.

The abnormality found on the POCR in all four of the adjournments was concerned a pneumonia (indicated as infiltrate in Table 2). All four patients were aged 74 or higher, had ASA scores of three or higher and had elevated C-reactive protein (CRP) levels and white blood cell (WBC) counts in their blood samples besides the clinical symptoms of fever, coughing, shortness of breath and abnormalities during auscultation. With a POCR costing €50, approximately €32,000 could have been saved during this study period if a POCR was only made in these cases.

Out of the adjourned operations, one patient died before surgery because of a very poor pre-hospital condition with preexisting terminal heart failure.

Out of the eight POCR’s showing signs of pulmonary edema (Table 2), there were either clear clinical symptoms in all patients, new ECG abnormalities or no clinical influence of the pulmonary edema on the perioperative care. The four POCR’s that showed signs of neoplasm’s were all preexistent. Eleven of the abnormalities classified as other in Table 2 were consolidations, densities or nodules of unknown origin of which two abnormalities lead to CT scans during hospital stay, where one lung carcinoma was found.

Many factors involving the treatment of hip fracture patients are of importance in minimizing the risk of complications and mortality during and after admission, but the additional value of POCR in hip fracture patients is limited, especially in patients with no clinical abnormalities prior to surgery. The POCR in our study showed abnormalities in 145 out of 642 (22.6%) patients with most of these being abnormalities matching with chronic diseases not requiring intervention. Only four out of these lead to an adjournment of the operation because of a chest infection. All four patients had clinical symptoms (fever, coughing, abnormal breath sounds during examination or elevated infection parameters) that supported the findings on the POCR. No significant differences in peri- and postoperative complications in relation with the result of the POCR were found.

Previous studies also suggest that the consequences of the abnormalities found on the POCR for scheduled operations are low, although only performed in elective patient groups. Guidelines of the European Society of Anesthesiology show that abnormal POCR only alter perioperative management in 3% out of the 23.1% of abnormal POCR’s [6]. Our study shows similar results and adds to the current evidence, being the first to describe the limited consequences of the POCR in a nonelective, fragile and specific surgical patient population.

The additional value of routine POCR is not only very limited but also associated with costs, potential delay of surgery because of unnecessary subsequent testing related to possible abnormal findings without any clinical relevance, causing worsening of the primary surgical condition and avoidable radiation risk for the patient [1]. An early intervention seems obvious since hip fractures lead to immobilization with many possible consequences. However, hip fracture patients often have significant comorbidities and a short delay to identify correctable ones in order to optimize their clinical condition is also logical. Mixed results are found when reviewing the ideal timing for surgery. An extensive review of the evidence for the ideal timing of surgery by Lewis et al suggests a trend for early operation within 12–48 h for ASA 1 and 2 patients [13]. Further delay over 48 h showed a rise in postoperative mortality. On the other hand, a delay up to five days for ASA 3 and 4 patients did not alter mortality. Though, a higher mortality rate was seen in patients of these categories when significant comorbidities were not corrected. However, the delay leaves the patient with significant pain and discomfort and the risk for pressure sores [13]. In our study a delay from 24 h onward in ASA 1, 2 (not significantly) and 3 patients resulted in more postoperative complications during admission, but did not alter mortality. There were no statistical differences showing an optimal window for ASA 4 patients. The outcome of POCR in general had no influence on the time till surgery.

The type of anesthesia is also an important factor. A recent study comparing RA versus GA in 328,540 matched surgical procedures showed that RA is associated with a 24% lower chance of respiratory complications, but did not alter 30-day mortality [14]. When comparing GA and RA in hip fracture patients, not only the incidence of postoperative respiratory complications decreases but also a 29% lower odds for mortality was seen in favor of the GA. Interestingly enough the advantages were only seen in patients with intertrochanteric fractures and not femoral neck fractures [15]. Our study also showed that patients undergoing GA were at higher risk for postoperative complications in general and especially respiratory complications. The differences based on fracture type were not found in our study.

Patients who are classified in a higher ASA group undergo GA and have a longer time till surgery independently have a higher chance of developing postoperative complications during admission. The POCR were not of any influence in these three factors.

There is always a chance of incidental findings on the POCR. One new incidental neoplasm (0.16%) was found during the study period.

The point should be raised that the retrospective design could have resulted in incomplete data collection. It was unknown how many patients were confused, which could influence the value of the preoperative screening because of difficult history taking. The follow-up of patients was limited to the time in hospital. The single center design does possibly limit the generalizability of this study. However, it can be reasoned that the geriatric population with a hip fracture does not differ between hospitals and therefore multicenter studies might not add more new evidence.

Some patients will need postoperative chest radiographs, and a possible benefit of the POCR is that it provides a comparative film. Some POCR’s will also act as a reassurance when there is doubt over the clinical condition of a patient. However, if the condition of the patient is doubtful or changes postoperatively, a chest radiograph is indicated without the need for a comparative film.

The interobserver variation in interpretation of the POCR has not been taken into account in this study. False negative results and false positive results can lead to per- and postoperative complications and unnecessary and potential harmful diagnostics and extended hospital stay, respectively. However, in this way the human error during daily practice is simulated.

The incidence of abnormal POCR rises with age [2]. The age of patients whose operation was adjourned because of an abnormal POCR was all above 70 years with ASA scores three or higher. Based on these data we cannot assess whether there are other groups of patients, besides symptomatic patients, who have an indication for a POCR since the greater part of all patients in this study are aged above 70 years with high ASA scores.

Most patients with abnormalities on the POCR that could have influenced perioperative care had either clinical symptoms or abnormal preoperative diagnostic tests. Even though this study had a retrospective design, it can be reasoned that performing selective POCR in case of clinical abnormalities can be a safe practice for patients with a (LET) hip fracture.

The value of the POCR in patients with a hip fracture is limited. Only in 0.6% of all performed POCR’s an abnormality leads to the adjournment of the operation. In these cases, the POCR matched the clinical findings. The known risk factors (GA, longer time till surgery and higher ASA classifications) for developing postoperative complications were not influenced by the POCR. The result of the POCR did not influence the rate of postoperative complications or mortality, and there were no cardiothoracic-related perioperative complications related to the results of the POCR. The preoperative screening in hip fracture patients is vital, though we think that the selective use of POCR in clinical abnormalities is safe and will reduce unnecessary delay and costs without compromising the quality of perioperative care.