This Article Abstract Résumé de cet Article Full Text (PDF) Additional Material Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Joo, H. S. Articles by Savoldelli, G. L. Search for Related Content PubMed PubMed Citation Articles by Joo, H. S. Articles by Savoldelli, G. L.

The value of screening preoperative chest x-rays: a systematic review

[La valeur des radiographies pulmonaires de dépistage préopératoire : une revue systématique]

Hwan S. Joo, MD FRCPC^*, Jean Wong, MD FRCPC, Viren N. Naik, MD Med FRCPC^* and Georges L. Savoldelli, MD^*

^* From the Departments of Anesthesia, St. Michael’s Hospital, and
the Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada.

Address correspondence to: Dr. Hwan Joo, Department of Anesthesia, St. Michael’s Hospital, 30 Bond Street, Toronto M5B 1W8, Canada. Phone: 416-864-5071; Fax: 416-864-6014; E-mail: jooh{at}smh.toronto.on.ca

	Abstract

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 
Purpose: Chest x-ray (CXR) is the most frequently ordered radiological test in Canada. Despite published guidelines, variable policies exist amongst different hospitals for ordering of preoperative CXRs. The purpose of this study was to systematically review the literature on the value of screening CXRs and establish evidence to support guidelines for the use of preoperative screening CXRs.

Source: Medline and Embase were searched under set terms for all English language articles published during 1966–2004. All eligible studies were reviewed and data were extracted individually by two authors. Of the 513 articles identified, 14 studies met both inclusion and exclusion criteria.

Principal findings: The quality of published evidence was modest as only six of the studies were rated as fair and eight as poor. Of the reported studies, diagnostic yield increased with age. However, most of the abnormalities consisted of chronic disorders such as cardiomegaly and chronic obstructive pulmonary disease (up to 65%). The rate of subsequent investigations was highly variable (4–47%). When further investigations were performed, the proportion of patients who had a change in management was low (10% of investigated patients). Postoperative pulmonary complications were also similar between patients who had preoperative CXRs (12.8%) and patients who did not (16%).

Conclusion: An association between preoperative screening CXRs and decrease in morbidity or mortality could not be established. As the prevalence of CXR abnormalities is low in patients under the age of 70, there is fair evidence that routine CXRs should not be performed for patients in this age group without risk factors. For patients over 70, there is insufficient evidence for or against performance of routine CXRs. The current recommendation from the Guidelines Association Committee that routine CXRs should not be performed for patients over 70 without risk factors is supported by this study.

	Introduction

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 
HISTORICALLY, the practice of ordering chest radiographs routinely for all hospital admissions and preoperatively began during World War II to identify patients with asymptomatic pulmonary tuberculosis.¹ However, with the declining prevalence of this disease and the development of better screening tests, mass radiologic screening programs have stopped. Moreover, in non-endemic populations, routine chest x-rays (CXRs) have not been shown to be efficacious for detecting asymptomatic tuberculosis or cancer.²

The clinical practice guidelines published by the Guidelines Advisory Committee (GAC), and supported by the Ministry of Health and Long Term Care in Ontario and the Ontario Medical Association, advocate that routine preoperative CXR is not warranted (www.gacguidelines.ca). Accordingly, the Canadian Anesthesiologists’ Society suggests only performing preoperative CXR based upon specific indications after conducting a history and physical examination (http://www.cas.ca/members/sign_in/guidelines/preanesthetic/). Despite these recommendations, policies remain variable from one hospital to another regarding preoperative CXRs. Routine preoperative CXRs have been ordered in some hospitals based solely upon age criteria.³

CXRs are the most frequently ordered radiological test in the United States, accounting for up to 45% of all diagnostic radiological procedures at a cost of $1.5 billion.^1,4 Although data for Canada are unavailable, an estimation that one-tenth of the number of CXRs ordered in the United States seems appropriate. The purpose of this review is to perform a systematic qualitative assessment regarding the value of screening CXRs and establish evidence to support current guidelines. By assessing the prevalence of respiratory abnormalities, their impact on anesthetic or surgical management, and the morbidity and mortality associated with the abnormalities, we aim to determine if preoperative CXRs are helpful in any specific patient population.

	Methods

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 
Medline was searched for all English language articles published between 1966 and April 2004. The medical subject headings used for the Medline search were "radiography, thoracic", "preoperative care", and "postoperative complications". Embase was searched for articles in English relating to preoperative chest radiography published between 1980 and March 2004, using the descriptors "thorax radiography", "preoperative period", "postoperative period". The Cochrane database of clinical trials was also searched for articles relating to preoperative chest radiography. Scanning the bibliography of review articles and all included articles identified additional articles.

Studies were included if they were prospective or retrospective, and investigated the utilization of preoperative chest radiographs on adult patients undergoing non-cardiac and non-thoracic surgery in developed countries. Excluded were studies of pediatric populations, non-English articles, and articles originating from non-first world countries due to the high prevalence of tuberculosis in those countries.⁵

All eligible studies were reviewed and data extraction was performed individually by two authors (J.W. and H.J.) independently. The quality of the studies was assessed in accordance with the Canadian Task Force on Preventive Health Care’s guidelines for grading quality of published evidence. Studies were graded as good, fair or poor (Appendix I, available as Additional Material at www.cja-jca.org). Differences between reviewers were resolved by consensus after review of the original studies. Grading of recommendations was based upon the grading system established by the Canadian Task Force on Preventive Health Care (Appendix II, available as Additional Material at www.cja-jca.org).

The impact of abnormal CXRs was assessed by noting the percentage of patients with abnormal CXRs who had: 1) further investigations or optimization of identified medical problems due to CXR abnormalities; 2) abnormal CXRs leading to a change in anesthetic management; 3) abnormal CXRs leading to a change in surgical management; and 4) abnormal CXRs leading to perioperative complications.

	Results

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 
Of the 513 articles identified, 14 studies met both inclusion and exclusion criteria^6–19 (Appendix III, available as Additional Material at www.cja-jca.org). All studies were non-controlled and non-randomized. Only three of the 14 studies were prospective and rated fair (II-2 to II-3).^7,14,19 Three of the 11 retrospective studies were rated fair (II-3)^6,16,18 and the remaining eight studies were rated poor (III).^8–13,15,17

Diagnostic yield of preoperative chest radiographs
Diagnostic yield was defined as:

Eight studies stratified the diagnostic yield of preoperative CXRs based upon the presence of risk factors including history of cardiovascular, respiratory, and other medical diseases. All of the studies demonstrated an increase in the diagnostic yield of CXRs in patients who had risk factors (Appendix III). However, the risk factors were subjective and varied. They were often classified as "coexisting bronchopulmonary or cardiac conditions, abnormal clinical cardiopulmonary findings or ASA > II".^7,14

Only two studies reported whether the abnormalities were new or unexpected findings. The first study by Bouillot⁷ found that 56% of the CXR abnormalities were chronic, and the second study by Wiencek found that 64% of the CXR abnormalities were previously known.¹⁹ In fact, most of the abnormalities consisted of chronic disorders such as cardiomegaly (15–65%) or chronic obstructive pulmonary disease (COPD); (7–30%). Four studies reported the acute processes of pulmonary congestion or congestive heart failure (CHF); (6–25%).^6,7,17,18

Nine studies stratified the diagnostic yield of preoperative CXRs in adult non-cardiothoracic surgical patients by age.^{6–9,12–14,17,18} The diagnostic yield of preoperative CXRs in patients under the age of 50 was low, ranging from 3 to 16% (Table I). The diagnostic yield reported for patients between the ages of 51 to 60 was variable and ranged from < 10% to as high as 59% in one study. The diagnostic yield increased to between 13 to 58% for patients between the ages of 61 to 70, and the diagnostic yield was consistently high, ranging between 47 to 61% for patients > 70 yr of age, except for one study by Umbach¹⁸ with a 27 to 33% yield.

Anesthetic management
Five studies reported whether changes in anesthetic management occurred in patients with CXR abnormalities.^7,8,14,15,19 One large prospective study by Bouillot reported that only 0.5% of patients had a change in anesthetic or surgical management based upon the results of abnormal CXRs.⁷ This prospective study explicitly defined, a priori, that significant changes in anesthetic management would include changes in anesthetic drugs, type of anesthetic, and postponement of surgery.

In one large but dated retrospective study by the Royal College of Radiologists, there was no difference in the percentage of patients with normal or abnormal CXRs who received inhalational anesthesia.¹¹ In two retrospective studies by Tape and Umbach, CXR abnormalities led to changes in anesthetic management in 5%, and 5.8% of patients respectively.^16,18 The type of change in management, such as more invasive monitoring, change in type of anesthesia or change in postoperative care was not reported.

Surgical management
Only one study reported cancelling surgery as a result of the CXR finding in 2% of patients.¹⁹ A study by Gagner reported delaying surgery in 1.3% of patients based on a preoperative CXR indicating pneumonia.⁸ Changes in surgical management occurred in 1% of patients with no risk factors, and 4% of patients with risk factors in Sommerville’s study.¹⁵

Perioperative complications
In a study by Boghosian, perioperative complications occurred more frequently in patients > 70 (17%), vs patients between 60 to 70, (4%).⁶ A study by Mendelson reported that a preoperative chest radiograph was useful for comparison in 51% of patients having postoperative chest radiographs.¹⁰ However, the clinical benefits of such a comparison were not reported. The usefulness of comparing an upright posterior-anterior and lateral preoperative CXR to a supine, anterior-posterior CXR has been questioned.²⁰ The Royal College of Radiologists reported that postoperative pulmonary complications occurred in 12.8% of patients who had a preoperative CXRs and 16% of patients who did not have a preoperative CXRs.¹¹ They did not specifically examine whether patients with abnormal preoperative CXRs had a higher incidence of postoperative complications. Two studies reported no difference in perioperative complications between patients with abnormal or normal chest radiographs; 13 vs 9%, and 3 vs 2%, respectively.^16,18

The only common diagnosis on CXR that may affect perioperative outcome is CHF.²¹ However, the four studies that reported CHF did not state whether it was an unexpected finding and whether the CXR findings had a major role in affecting management. Presently, routine CXR screening for CHF in asymptomatic patients is not part of clinical practice.

	Discussion

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 
Our review suggests that the diagnostic yield of a preoperative CXR increased with age. However, most of the abnormalities reflected chronic disorders. The rate of subsequent investigations was highly variable, and when further investigations were performed, the proportion of patients who had a change in management was low. Finally, the incidence of postoperative pulmonary complications appeared to be similar between patients who had preoperative CXRs and patients who did not.

The use of routine CXR is associated with risks and costs. The lifetime risk of cancer death from routine CXR has also been estimated to be about 1.2 per 100,000.²² As well, a false positive result, and subsequent delay of surgery caused by additional unnecessary investigations may potentially lead to patient harm and result in worsening of the primary surgical condition.

Most studies in this analysis had methodological problems that limited the interpretation of the results and strength of their conclusions. Many studies did not adequately report outcomes; thus the impact of abnormal preoperative CXRs could not be determined. The reporting of postoperative morbidity and mortality was incomplete, follow-up periods were variable, and the types of complications were not discernable in many studies. Thus, the prognostic yield of abnormal preoperative CXRs could not be determined. The retrospective nature of the majority of the studies may have accounted for the lack of reporting further investigations, adverse effects, and changes in perioperative anesthetic and surgical management.

One of the major problems with assessing the accuracy of the CXR is the lack of a suitable gold standard for the many abnormalities detectable with this test. Interpreter variability is likely to increase the inaccuracy of the results. There may be an overestimation of the usefulness of CXRs to pick up disease as the sensitivity and specificity for tuberculosis was reported as 75 and 98%, 91 and 96% for COPD and 74 and 88% for CHF.²³ Therefore the diagnostic yield may have errors which may underestimate or overestimate the true incidence of the abnormality.

None of the studies reported whether the radiologists were blinded to the clinical history or physical findings of the patients, thus, radiologists’ interpretation of CXRs may be affected by bias and the diagnostic yield reported may be higher than what would be expected in clinical practice.²⁴ The retrospective nature of most of the studies makes blinding of radiologists unlikely.

To be an effective screening examination, the test has to correctly identify the patients at risk for future mortality or morbidity and have an effective treatment available.²⁵ The test should have a high sensitivity and a low incidence of adverse or side effects. The test should also lead to a decrease in mortality or morbidity either directly or indirectly.²⁶ In this review of the available literature on the effectiveness of routine preoperative CXRs, these qualities were not adequately fulfilled by the test.

Although none of the studies reviewed were randomized controlled trials, this type of study design may not be necessary if it could be demonstrated that abnormal CXRs resulted in higher morbidity/mortality, further investigations, or changes in perioperative anesthetic or surgical management. However, only the prevalence of abnormal CXRs was measured routinely, and perioperative mortality and morbidity was not well reported in the majority of studies. Abnormal CXRs led to more investigations in some studies, however, it cannot be discerned whether the patients actually benefited from more investigations.

There is evidence in four studies of fair quality,^6,7,14,18 to support the recommendation that routine preoperative CXRs be eliminated in asymptomatic patients who are less than 70 yr of age (Grade D recommendation, Appendix II). We make this recommendation based upon the low incidence of abnormal CXRs in this patient population, and the fact that abnormal CXR findings have not been shown to help in decreasing perioperative morbidity in any of the 14 studies reviewed. Even when abnormalities were present, the majority of abnormalities consisted of chronic conditions such as cardiomegaly and COPD. Additionally, the incidence of perioperative complications in these patients is low.

The above recommendation does not mean those patients above the age of 70 should have routine preoperative CXRs, as detection of asymptomatic abnormalities on CXR has not been shown to correlate with perioperative outcome. Without firm evidence from the literature (Grade C, Appendix II), we defer to the published GAC guidelines that suggest eliminating routine preoperative CXRs without a clinical indication, regardless of age.

	Summary and recommendations

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 

1. Prevalence of CXR abnormalities increases with age and risk factors. Risk factors include cardiopulmonary comorbidities and severe systemic illnesses.
   
2. Most abnormalities on preoperative CXR are chronic, with the majority of the abnormalities being cardiomegaly and COPD.
   
3. Most CXR abnormalities, with the possible exception of acute heart failure, have not been shown to affect anesthetic or surgical management, or perioperative outcome.
   
4. Based upon the best available current evidence, routine preoperative CXRs should not be performed in asymptomatic patients, regardless of age.
   

TABLE II Summary of impact on perioperative management and complications

Author and year Further investigations due to CXR abnormalities Change in anesthetic management Change in surgical management Perioperative complications Boghosian 19876 Overall : 9.6% Ages 60–70: 4% > 70: 17% Bouillot 19967 Change in anesthetic/surgical management only because of abnormal CXR findings No risk factors: 0.1% Mild risk factors: 0.3% Moderate risk factors: 1% High risk factors: 1% Overall: 0.5% CXRs helpful for diagnosis of complication No risk factors: 2% Mild risk factors: 7% Moderate risk factors: 10% High risk factors: 14% Overall: 5% Gagner 19908 1.3% Mendelson 198710 Useful for comparison with 51% of patients having postoperative CXRs Royal College of Radiologists 197911 96.7% of patients with normal CXRs vs 96.1% with abnormal CXRs had inhalation anesthesia Patients with preoperative CXRs had 12.8% incidence of complication vs 16% for those who had no preoperative CXRs Silvestri 199914 2.4% 3.7% Sommerville 199215 1% (no risk factors) 4% (risk factors) Tape 198816 5% 13% vs 9% (abnormal vs normal CXRs) Umbach 198818 47% 5.8% 3% vs 2% (abnormal vs normal CXRs) Wiencek 198719 4% 2% CXR = chest x-rays.

TABLE III Summary types of abnormalities seen on CXRs

Cardiomegaly Chronic obstructive pulmonary disease Pulmonary congestion Boghosian 19876 15% 27% 7% Bouillot 19967 31% 6% McCleane 19899 25% 30% Mendelson 198710 40% 13% Rees 197612 54% 19% Silvestri 199914 34% Tornebrandt 198217 65% 25% Umbach 198818 51% 7% 10% Wiencek 198719 42%

	Acknowledgments

 
We thank Dr. Valerie Palda and Dr. Denice Feig of The Canadian Task Force on Preventive Health Care (www.ctfphc.org) for their help in guiding the authors with this manuscript.

	Footnotes

 
Reprints will not be available from the author.

Assessed August 5, 2004. Accepted for publication November 18, 2004. Revision accepted February 17, 2005.

	References

TOP Abstract Introduction Methods Results Discussion Summary and recommendations References

 
1 Sagel SS, Evens RG, Forrest JV, Bramson RT. Efficacy of routine screening and lateral chest radiographs in a hospital-based population. N Engl J Med 1974; 291: 1001–4.

2 Anonymous. A rational approach to radiodiagnostic investigations. World Health Organ Tech Rep Ser 1983; 689: 1–49.[Medline]

3 McKee RF, Scott EM. The value of routine preoperative investigations. Ann R Coll Surg Engl 1987; 69: 160–2.[Medline]

4 Hubbell FA, Greenfield S, Tyler JL, Chetty K, Wyle FA. The impact of routine admission chest x-ray films on patient care. N Engl J Med 1985; 312: 209–13.[Abstract]

5 Anonymous. WHO-Expert Committe on Tuberculosis, Ninth Report. Technical Report Series No. 552; WHO, Geneva, 1974 (German). Z Erkr Atmungsorgane 1975; 143: 162–93.[Medline]

6 Boghosian SG, Mooradian AD. Usefulness of routine preoperative chest roentgenograms in elderly patients. J Am Geriatr Soc 1987; 35: 142–6.[Medline]

7 Bouillot JL, Fingerhut A, Paquet JC, Hay JM, Coggia M. Are routine preoperative chest radiographs useful in general surgery? A prospective, multicentre study in 3959 patients. Association des Chirurgiens de l’Assistance Publique pour les Evaluations médicales. Eur J Surg 1996; 162: 597–604.[Medline]

8 Gagner M, Chiasson A. Preoperative chest x-ray films in elective surgery: a valid screening tool. Can J Surg 1990; 33: 271–4.[Medline]

9 McCleane GJ. Routine preoperative chest x-rays. Ir J Med Sci 1989; 158: 67–8.[Medline]

10 Mendelson DS, Khilnani N, Wagner LD, Rabinowitz JG. Preoperative chest radiography: value as a baseline examination for comparison. Radiology 1987; 165: 341–3.[Abstract/Free Full Text]

11 Anonymous. Preoperative chest radiology. National study by the Royal College of Radiologists. Lancet 1979; 2: 83–6.[Medline]

12 Rees AM, Roberts CJ, Bligh AS, Evans KT. Routine preoperative chest radiography in non-cardiopulmonary surgery. Br Med J 1976; 1: 1333–5.

13 Rucker L, Frye EB, Staten MA. Usefulness of screening chest roentgenograms in preoperative patients. JAMA 1983; 250: 3209–11.[Abstract]

14 Silvestri L, Maffessanti M, Gregori D, Berlot G, Gullo A. Usefulness of routine pre-operative chest radiography for anaesthetic management: a prospective multicentre pilot study. Eur J Anaesthesiol 1999; 16: 749–60.[Medline]

15 Sommerville TE, Murray WB. Information yield from routine pre-operative chest radiography and electrocardiography. S Afr Med J 1992; 81: 190–6.[Medline]

16 Tape TG, Mushlin AI. How useful are routine chest x-rays of preoperative patients at risk for postoperative chest disease? J Gen Intern Med 1988; 3: 15–20.[Medline]

17 Tornebrandt K, Fletcher R. Pre-operative chest x-rays in elderly patients. Anaesthesia 1982; 37: 901–2.[Medline]

18 Umbach GE, Zubek S, Deck HJ, Buhl R, Bender HG, Jungblut RM. The value of preoperative chest X-rays in gynecological patients. Arch Gynecol Obstet 1988; 243: 179–85.[Medline]

19 Wiencek RG, Weaver DW, Bouwman DL, Sachs RJ. Usefulness of selective preoperative chest x-ray films. A prospective study. Am Surg 1987; 53: 396–8.[Medline]

20 Milne EN, Burnett K, Aufrichtig D, McMillan J, Imray TJ. Assessment of cardiac size on portable chest films. J Thorac Imaging 1988; 3: 64–72.[Medline]

21 Goldman L, Caldera DL, Southwick FS, et al. Cardiac risk factors and complications in non-cardiac surgery. Medicine (Baltimore) 1978; 57: 357–70.[Medline]

22 Anonymous. Report from the Conseil d’Evaluation des Technologies de la Santé (CETS). Mandate of the conseil. Int J Technol Assess Health Care 1992; 8: 366.[Medline]

23 Tape TG, Mushlin AI. The utility of routine chest radiographs. Ann Intern Med 1986; 104: 663–70.

24 Davies HT. Bias in treatment trials. Hosp Med 1999; 60: 599–601.[Medline]

25 Roizen MF. More preoperative assessment by physicians and less by laboratory tests (Editorial). N Engl J Med 2000; 342: 204–5.[Free Full Text]

26 Eagle KA, Brundage BH, Chaitman BR, et al. Guidelines for perioperative cardiovascular evaluation for noncardiac surgery. Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). J Am Coll Cardiol 1996; 27: 910–48.[Medline]

This article has been cited by other articles:

				S. R. Bapoje, J. F. Whitaker, T. Schulz, E. S. Chu, and R. K. Albert Preoperative Evaluation of the Patient With Pulmonary Disease Chest, November 1, 2007; 132(5): 1637 - 1645. [Abstract] [Full Text] [PDF]

				G. W. Smetana A 68-Year-Old Man With COPD Contemplating Colon Cancer Surgery JAMA, May 16, 2007; 297(19): 2121 - 2130. [Abstract] [Full Text] [PDF]

				A. Hamilton Preoperative testing. Can J Anesth, March 1, 2006; 53(3): 326 - 327. [Full Text] [PDF]

				G. L. Bryson Has preoperative testing become a habit?/L'administration de tests preoperatoires est-elle devenue une habitude? Can J Anesth, June 1, 2005; 52(6): 557 - 561. [Full Text] [PDF]

This Article Abstract Résumé de cet Article Full Text (PDF) Additional Material Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Joo, H. S. Articles by Savoldelli, G. L. Search for Related Content PubMed PubMed Citation Articles by Joo, H. S. Articles by Savoldelli, G. L.