The Outcome of the Follow-Up of Consolidations on Chest Radiographs in a Maltese Population, Presenting from the Community, Aged 50 or over – a Retrospective Study.
Abstract
Background: The British Thoracic Society (BTS) guidelines for community-acquired pneumonia (CAP) suggest a repeat chest radiograph 6 weeks after treatment for patients over the age of 50 to screen for lung malignancy. The benefit of thispractice is not well determined.
Method: We conducted a retrospective study involving patients from the community over 50 years old with consolidations on chest radiography. These patients presented in Mater Dei Hospital, Gozo General Hospital and Maltese Health Centres during the months of January 2013-2017 and August 2013-2016.
The occurrence of follow-up imaging and subsequent diagnosis of lung malignancy was documented. All chest radiographs were reviewed by a radiologist.
Results: 402 patients met our inclusion criteria. Follow-up imaging was done in 214 patients (53.2%) within 12 weeks. There was no statistical significance in the follow-up rates when matched for the presenting month, whether radiologists recommended repeat imaging, whether patients were admitted to hospital, and for the patients’ age and gender.
The diagnostic yield of lung malignancy was 1.74% (7 patients) within 12 weeks with all malignancies being at an advanced stage at diagnosis (lowest stage being IIIA) when detected. All seven patients had a smoking history.
Conclusion: 53.2% of community-acquired pneumonia patients over the age of 50 had follow-up imaging within 12 weeks. No clinical variables explaining this low rate could be identified.
This practice results in a low diagnostic yield. Moreover, the diagnosis of lung malignancy is achieved at an advanced stage, making it a poor screening tool.
Scheduling repeat imaging in patients diagnosed with community acquired pneumonias (CAPs) with a consolidation on radiographs is routine practice for many a physician. Lung malignancies can have similar radiological features to that of a consolidation, while an airspace shadow caused by an infection can easily mask an underlying neoplasm. It is on this trail of thought that repeat imaging is routinely done within the first 2 to 3 months.
The 2009 British Thoracic Society (BTS) guidelines1 on Community Acquired Pneumonias mention that repeat chest radiography should be performed 6 weeks after the initial chest X-ray in patients over the age of 50 or those with a smoking history. Similarly, the 2005 American College of Chest Physicians guidelines2 suggest follow-up radiography after approximately 8 weeks.
On the other hand, the 2007 Infectious Disease Society of America and the American Thoracic Society consensus guidelines3 do not mention any follow-up chest radiography in the management of CAPs.
More detailed guidelines on follow-up imaging are conspicuous by their absence. Consequently, physicians are left with their clinical judgment and patchy guidelines when they need to decide who the patients meriting future chest imaging are.
To address this void in the management of community acquired pneumonias, we have embarked on this retrospective study where our aims were twofold. Firstly, are patients over the age of 50 diagnosed with a community acquired pneumonia really being followed-up, as suggested by the BTS guidelines, and if not, why? Secondly, is this practice really feasible and efficient? Using this strategy, what is the diagnostic yield of lung malignancy and is the diagnosis being achieved early enough?
Methodology
In this retrospective study, we considered patients 50 years old or over who presented to Mater Dei Hospital, Gozo General Hospital and Maltese Health Centres with radiological findings of a pneumonia.
The age 50 was chosen since the The 2009 British Thoracic Society (BTS) guidelines1 on Community Acquired Pneumonias actually consider patients over the age of 50 to be at high risk for lung malignancy when recommending repeat chest imaging.
Patients were recruited by searching for the keywords “consolidation”, “pneumonia” and “opacification” in the radiologist reports for all the radiographs done in the months of January during the years 2013, 2014, 2015, 2016 and 2017 and the months of August in the years 2013, 2014, 2015 and 2016.
These reports were obtained using the local Picture Archiving and Communication System (PACS) software used in all Maltese state hospitals and health centres.
The patients whose reports had one or more of these keywords were then sieved through according to the inclusion and exclusion criteria as mentioned in Tables 1 and 2.
The consolidation must have been present only on chest radiographs. Thus, patients who had other modalities of chest imaging for up to one week after the presenting chest radiograph were excluded. Since this study focuses specifically on community acquired pneumonias, patients who were admitted for any reason in hospital during the three weeks prior to the chest radiograph were excluded. Patients with active malignancy were also excluded.
A radiologist higher specialist trainee reviewed all the chest radiographs which were reported to have a consolidation. This helped ensure that only chest radiographs with clear signs of consolidations were included in this study.
The final number of patients who satisfied all the aforementioned was 402 (n).
The patients’ demographics were documented along with the mortality and whether or not follow-up imaging was done or not. The initial chest radiograph reports were analysed to see whether the reporting radiologist had recommended follow-up imaging and consequently whether this affected the rate of follow-up or not. Other variables that may have affected rate of follow-up such as whether the patient was admitted and whether the consolidations were unilateral or bilateral were also documented.
All follow-up imaging done from one week until 12 weeks after the initial chest radiograph was scrutinized. The data collected was analysed using Statistical Package for Social Sciences (SPSS) software.
Where relevant, the histology, if available, and stage of the lung malignancy were noted. The 8th edition TNM classification4 was used to assess staging.
Results
As mentioned previously, a total of 402 patients met our inclusion criteria. 207 were male (51.5%) and the age ranged from 50 till 99 years. The mean age was 74.18 years (S.D. ±11.8 years) with the median being 76 years.
Follow-up imaging was performed in 214 patients within 12 weeks after the initial chest radiograph. This implies that 214 patients had chest imaging done between 1 week and 12 weeks after initial presentation (as documented in Table 3). Chest radiographs done on admitted patients who were in hospital for at least 24 hours were excluded.
316 patients of the cohort of 402 patients (78.6%) required hospital admission after the initial chest radiograph. Repeat imaging after treatment of the pneumonia was suggested by a radiologist in 130 chest radiographs, i.e. in 32.3% of the CAP population. Refer to Tables 3 and 4 for an overview of the data collected.
Different follow-up rates within 12 weeks were compared according to gender, the month of presentation and whether follow-up was recommended by radiologist or not (Figure 1), if the consolidations was bilateral and whether the patient required admission by using a chi square statistical test for each variable.
{*figure 1*}
Taking a p-value of <0.05 as being statistically significant, it was noted that no significant difference was detected among the follow-up rates when comparing these different predictors (Table 5). A paired t-test was done to compare age and follow-up rates. Once more, no statistical significance was noted (Table 6).
{*table 5*}
{*table 6*}
Additionally, logistic regression testing was done to assess any interaction among all the aforementioned predictors that may affect the follow-up rate. Once again, no significant difference in follow-up rate was detected.
A total of 58 patients (i.e. 27.1% of all patients followed up) had non-resolving radiological findings on repeat imaging. 38 of these were eventually diagnosed with benign conditions while 13 did not have further imaging done.
The remaining 7 patients had non-resolving radiological findings that led to the diagnosis of lung malignancy. This means that 3.27% of patients followed up were diagnosed with lung malignancy, while the diagnostic yield (i.e. the number of patients diagnosed with lung cancer compared to the total number of patients with a community-acquired pneumonia) was 1.74%.
The diagnostic yield in the months of January was 1.48% while that in the months of August was similar at 2.27%. Using proportion testing, this difference was not significant (p-value: 0.570).
Diagnosis of lung cancer was achieved at a relatively late stage in all cases; the most favourable stage was IIIA (as documented in Table 7).4
{*table 7*}
In addition to the aforementioned data, when including all radiology studies from one week after presentation up until 12 months after, a total of 320 patients had chest imaging done, i.e. 79.6% of the total cohort. In this scenario, 8 patients had unresolved lesions that led to a diagnosis of lung malignancy, i.e. only one patient was diagnosed with lung malignancy after not being followed up within 12 weeks. All 8 patients had a smoking history and their age ranged from 61 till 80 years (as shown in Figure 2). 7 of these patients were males. A cytology and/or histology confirmation of malignancy was achieved in 7 patients; one patient unfortunately passed away before this was achieved with his diagnosis only being done using radiological means. Three cases were adenocarcinoma, two squamous cell carcinoma and two small cell carcinoma. All were reported to be likely bronchial in origin. Table 8 gives a detailed account of these 8 patients.
{*figure 2*}
Moreover, similar to the rest of the cohort, repeat imaging had only been suggested in 4 of the 8 patients who were diagnosed with lung malignancy within one year.
Discussion
The most striking consideration from this study is that lung malignancy was diagnosed at a late stage during radiological follow-up of community acquired pneumonias. Using such a practice as a screening tool for lung malignancy is clearly inadequate from our data. None of the patients diagnosed with lung cancer on follow-up could reasonably be offered curative surgery as the most favourable lung cancer stage was IIIA. To our knowledge, no recent studies have been done delving into the actual stage of the lung cancer on diagnosis when following-up community acquired pneumonias.
Secondly, and equally as noteworthy, is the low diagnostic yield of lung malignancy on follow-up of community acquired pneumonias. Only 1.74% of patients who had follow-up chest imaging within 12 weeks were diagnosed with lung malignancy. This diagnostic yield did not vary significantly from the months of January to August.
53.2% of CAP patients were followed up with repeat imaging within 12 weeks, a figure that is similar to that in other studies.5-7 This seems to suggest that multiple centres do not feel the need to adhere to guidelines such as those of the British Thoracic Society in each and every patient. The reason for this is likely multifactorial. Follow-up rate was however not affected by the presenting month, age and gender of the patients and neither by whether or not the patients were admitted, nor whether the radiologist suggested repeat imaging when reporting the first chest radiograph.
The fact that only one case of lung malignancy was diagnosed within one year in those patients who were followed-up within 12 weeks seems to suggest that physicians are picking up clinical clues that help choose which patients are most likely to benefit from repeat imaging. This study was unable to identify any such indicators.
Larger, prospective studies looking into multiple patient co-morbidities and demographics may help identify other, more specific clinical predictors that can better guide physicians to decide regarding follow-up imaging.
Limitations
This study had various limitations that one must point out. Firstly, the study was carried out retrospectively in a relatively small cohort. Similar studies have been carried out, however they are sporadic and thus, it was difficult to compare results. The clinical features on presentation, patient co-morbidities (apart from the presence of an active malignancy) and serology results that aid in the diagnosis of a pneumonia were not included in this study making it heavily reliant on radiologic findings.
Moreover, it was impossible to obtain proper smoking histories and detailed documentation from the patients’ medical notes when they presented with the pneumonia. The reasons for this were several including the fact that a substantial cohort of patients have since passed away making access to their medical notes very difficult. Documentation outside Mater Dei Hospital, especially in the Health Centres, is also very sparse and not easily accessible.
Consolidations were identified only by three keywords on the radiograph report, thus the study was dependent on the report of the initial imaging. Only PACS software used in state-funded hospitals and health centres was analysed, meaning that follow-up imaging that could have potentially been done in the private sector or overseas was missed. This was mitigated to a certain extent by excluding patients without a Maltese fixed address.
Conclusion
This study shows that, similar to other studies done in different centres, the follow-up rate locally of consolidations on chest radiographs in patients presenting from the community is low (53.2%). Why this is so is still unclear.
The diagnostic yield of lung malignancies on follow-up within 12 weeks in patients over 50 years of age is just 1.74%. When analysing all chest imaging done within one year in all the patients who fitted our inclusion and exclusion criteria, lung malignancy was diagnosed in 1.99%.
When lung malignancy is detected, the stage was always noted to be advanced, and hence inoperable and with a poor prognosis. Following up consolidations on chest radiographs in community acquired pneumonias is a poor screening tool.
Summary Box
1. The follow-up rate of chest radiographs in patients presenting with a community-acquired pneumonia is 53.2% within 12 weeks in state-run hospitals and health centres in the Maltese Islands.
2. Lung malignancy is diagnosed at a late stage when following up community-acquired pneumonias.
3. The diagnostic yield of lung malignancy when following up community-acquired pneumonias within 12 weeks is 1.74%.
Acknowledgments
We would like to particularly thank Prof Josef Lauri for his invaluable guidance during the statistical analysis of our data.
Disclosure
The authors declare no conflict of interest.
References
Lim WS, Baudouin SV, George RC, Hill AT, Jamieson C, Le Jeune I et al. Pneumonia Guidelines Committee of the BTS Standards of Care Committee. BTS guidelines for the management of community acquired pneumonia in adults: update 2009. Thorax. 2009; 64(suppl 3): iii1-iii55.
Ramsdell J, Narsavage GL, Fink JB; American College of Chest Physicians’ Home care Network Working Group. Management of community-acquired pneumonia in the home: an American College of Chest Physicians clinical position statement. Chest. 2005; 127(5): 1752-1763.
Mandel LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC et al; Infectious Diseases Society of America; American Thoracic Society. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007; 44(suppl 2): S27-S72.
Detterbeck FC, Boffa DJ, Kim AW, Tanoue LT. The Eight Edition Lung Cancer Stage Classification. Chest. 2017; 151(1): 193-203.
Tang KL, Eurich DT, Minhas-Sandhu JK, Marrie TJ, Majumdar SR. Incidence, Correlates, and Chest Radiographic Yield of New Lung Cancer Diagnosis in 3398 Patients with Pneumonia. Arch Intern Med 2011;171(13): 1193-1198.
Little BP, Gilman MD, Humphrey KL, Alkasab TK, Gibbons FK, Shepard JO et al. Outcome of Recommentaions for Radiographic Follow-Up of Pneumonia on Outpatient Chest Radiography. AJR 2014; 202: 54-59.
Macdonald C, Jayathissa S, Leadbetter M. Is post-pneumonia chest X-ray for lung malignancy useful? Results of an audit of current practice. Intern Med J 2015; 45(3): 329-334.
![]()
Table 1:Inclusion criteria for patients to be eligible to form part of the study cohort
| Inclusion Criteria |
| Patients whose chest radiographs’ reports contained the search words “consolidation”, “opacification” and/or “pneumonia” |
| Age: 50 years or over |
| Consolidation must be seen on chest radiograph |
| Patients presenting in the months of January 2013-2017 |
| Patients presenting in the months of August 2013-2016 |
| Chest X-rays performed in Mater Dei Hospital, Gozo General Hospital and Maltese Health Centres |
Table 2:Exclusion criteria preventing patients from being eligible for the study cohort
| Exclusion Criteria |
| Patients under the age of 50 |
| Patients who had chest imaging other than chest radiographs done in the first week after the initial chest X-ray |
| Patients deceased within 12 weeks of initial chest radiographs |
| Hospital Acquired Pneumonias, i.e. patients who had been admitted in hospital at any stage during the three weeks prior to presentation |
| Chest radiographs that were never formally reported |
| Lesion described on chest x-ray was already identified in previous imaging |
| Patients with pleural effusions that required drainage |
| Lung transplant patients |
| Patients with active pulmonary tuberculosis |
| Patients who did not have a fixed address in Malta (thus potentially making follow-up less likely) |
| Chest X-rays done over 24 hours after admission if no CXR was performed on admission |
| Presenting CXRs where the reporting radiologist recommended cross-sectional imaging and/or bronchoscopy and/or PET scans due to a high index of suspicion for a lung malignancy |
Table 3:Overview of the data gathered from all 402 patients with radiological signs of a lung consolidation and presenting with a community-acquired pneumonia from the months of January and August 2013-2016 and the month of January 2017.
| Number of Patients (n) | 402 |
| Number of Males | 207 |
| Patients over 80 years old | 143 |
| Patients 50-80 years olds | 259 |
| Patients requiring admission | 316 |
| Follow-up imaging done (within 12 weeks) | 214 |
| % Follow-up done imaging within 12 weeks | 53.2% |
| Follow-up Recommended by Radiologist | 130 |
| Follow-up done after Radiologist Recommendation (within 12 weeks) | 76 |
| % 12 week follow-up rate after Radiologist Recommendation | 58.5% |
| 12 week follow-up in patients over 80 years old | 73 |
| % 12-week follow up in patients over 80 years old | 51.0% |
| 12 week follow-up in patients 50-80 years old | 141 |
| % 12-week follow up in patients 50-80 years old | 54.4% |
| Number of males with a follow-up chest imaging within 12 weeks | 110 |
| %12-week follow up in males | 53.1% |
| Number of females with a follow-up chest imaging within 12 weeks | 104 |
| %12-week follow up in females | 53.3% |
| Lung malignancy cases diagnosed on follow-up of non-resolving lesions | 7 |
| % diagnostic yield on follow-up (i.e. number of patients diagnosed with lung malignancy compared to total number - n) | 1.74% |
| % of followed-up patients diagnosed with lung malignancy | 3.27% |
| Chest imaging done within 12 months | 320 |
| % of cases having chest imaging done within 12 months | 79.6% |
| Lung malignancy cases diagnosed within 12 months | 8 |
| % potential diagnostic yield (as identified by looking at all imaging done within 12 months) | 1.99% |
Table 4:Comparing community-acquired pneumonias’ follow-up in the months of August 2013-2016 with the months of January 2013-2017
| Januaries 2013-2017 | Augusts 2013-2016 | |
| Number of Patients (n) | 270 | 132 |
| Number of Males | 133 | 74 |
| Patients over 80 years old | 96 | 47 |
| Patients 50-80 years olds | 174 | 85 |
| Follow-up imaging done (within 12 weeks) | 147 | 67 |
| % Follow-up done imaging within 12 weeks | 54.4% | 50.8% |
| Patients requiring admission | 215 | 101 |
| Follow-up done in admitted patients (within 12 weeks) | 114 | 53 |
| % 12 week follow-up rate in admitted patients | 53.0% | 52.5% |
| Follow-up Recommended by Radiologist | 98 | 32 |
| Follow-up done after Radiologist Recommendation (within 12 weeks) | 59 | 17 |
| % 12 week follow-up rate after Radiologist Recommendation | 60.2% | 53.1% |
| 12 week follow-up in patients over 80 years old | 46 | 27 |
| % 12-week follow up in patients over 80 years old | 47.9% | 57.4% |
| 12 week follow-up in patients 50-80 years old | 101 | 40 |
| % 12-week follow up in patients 50-80 years old | 58.0% | 47.1% |
| Number of males with a follow-up chest imaging within 12 weeks | 72 | 38 |
| %12-week follow up in males | 54.1% | 51.4% |
| Number of females with a follow-up chest imaging within 12 weeks | 75 | 29 |
| %12-week follow up in females | 54.7% | 50.0% |
| Lung malignancy cases diagnosed on follow-up of non-resolving lesions | 4 | 3 |
| % diagnostic yield on follow-up (i.e. number of patients diagnosed with lung malignancy compared to total number - n) | 1.48% | 2.27% |
| % of followed-up patients diagnosed with lung malignancy | 2.72% | 4.48% |
| Chest imaging done within 12 months | 217 | 103 |
| % of cases having chest imaging done within 12 months | 80.4% | 78.0% |
| Lung malignancy cases diagnosed within 12 months | 5 | 3 |
| % potential diagnostic yield (as identified by looking at all imaging done within 12 months) | 1.85% | 2.27% |
Table 8:Detailed overview of the patients who were diagnosed with lung malignancy within 12 months after first chest radiograph. All patients had a smoking history.
| Patient | Month when initial chest radiograph was done | Gender | Age (years) | Number of days after first chest radiograph when repeat chest imaging was done | Imaging Modality used for first follow-up | Histology |
| 1 | January 2013 | Male | 73 | 37 days | CXR | Squamous Cell Carcinoma |
| 2 | August 2013 | Male | 75 | 15 days | CXR | Adenocarcinoma |
| 3 | August 2013 | Male | 75 | 7 days | CXR | Small Cell Carcinoma |
| 4 | January 2014 | Male | 79 | 60 days | CXR | Small Cell Carcinoma |
| 5 | January 2015 | Male | 69 | 9 days | CXR | Adenocarcinoma |
| 6 | August 2015 | Male | 61 | 14 days | CT | Not available |
| 7 | January 2017 | Male | 67 | 25 days | CXR | Adenocarcinoma |
| 8 | January 2017 | Female | 80 | 283 days | CXR | Squamous Cell Carcinoma |
Figure
Test image
Issue
Section
Original Articles
Published
09-12-2019
Keywords:
Community-acquired pneumonia, follow-up imaging, Lung Malignancy, Screening
License
The Editorial Board retains the copyright of all material published in the Malta Medical School Gazette. Any reprint in any form of any part will require permission from the Editorial Board. Material submitted to the Editorial Board will not be returned, unless specifically requested.
