Fax: (011) 31 243 54 01 66
Quality of integrated care for patients with nonsmall cell lung cancer
Variations and Determinants of Care
Article first published online: 27 AUG 2007
Copyright © 2007 American Cancer Society
Volume 110, Issue 8, pages 1782–1790, 15 October 2007
How to Cite
Ouwens, M. M. M. T. J., Hermens, R. R. P. M. G., Termeer, R. A. R., Vonk-Okhuijsen, S. Y., Tjan-Heijnen, V. C. G., Verhagen, A. F. T. M., Hulscher, M. M. E. J. L., Marres, H. A. M., Wollersheim, H. C. H. and Grol, R. P. T. M. (2007), Quality of integrated care for patients with nonsmall cell lung cancer. Cancer, 110: 1782–1790. doi: 10.1002/cncr.22986
- Issue published online: 19 SEP 2007
- Article first published online: 27 AUG 2007
- Manuscript Accepted: 25 JUN 2007
- Manuscript Revised: 21 MAY 2007
- Manuscript Received: 14 MAR 2007
- Radboud University Nijmegen Medical Centre, The Netherlands
- The East Comprehensive Cancer Centre, The Netherlands
- quality of health care;
- quality indicators;
- health care;
- patient-centered care
In the current study, the authors focused on determinants influencing the quality of care and variations in the actual quality of integrated care for patients with nonsmall cell lung cancer (NSCLC) to estimate whether there is room for improvement.
The authors tested the quality of integrated care for 276 NSCLC patients with 14 quality indicators of professional (4 indicators), organizational (3 indicators), and patient-oriented quality (7 indicators). Patient characteristics and actual care data were derived from medical record data, patient-oriented care was derived from patient questionnaires, and professional and hospital characteristics were derived from questionnaires for professionals. The performance measure was the proportion of patients to whom the indicator applied who had positive scores on the indicator. Multilevel logistic regression analysis determined the influence of patient, professional, and hospital characteristics on care.
With regard to professional quality, the proportions of patients who underwent fluorodeoxyglucose–positron emission tomography or cervical mediastinoscopy according to the guideline criteria were 88% and 84%, respectively. Only 50% of the biopsies were adequately obtained during mediastinoscopy, and in 3% of the patients with clinical stage III disease (based on the TNM classification) there was a search for brain metastases before the initiation of combination therapy. With regard to organizational quality, the diagnostic route of 79% of the patients was completed within 21 days; 51% of patients began therapy within 35 days and 57% were discussed during multidisciplinary consultation. All but 1 patient-oriented quality indicator scored ≤58%. Hospitals varied by ≥20% with regard to 11 of the 14 indicators. The patient-related determinants “stage of disease,” “age,” and “comorbidity” were found to influence the indicator scores the most.
The quality of integrated care (especially patient-oriented care) for NSCLC patients needs improvement. Patient characteristics appear to influence performance more than professional or hospital characteristics. Cancer 2007. © 2007 American Cancer Society.
Lung cancer is the leading cause of cancer death in Europe and the U.S., killing more patients than breast, colon, and prostate cancers combined.1 Lung cancers are classified into 2 main categories: small cell lung cancer, which accounts for approximately 15% of the cases, and nonsmall cell lung cancer (NSCLC), which accounts for the other 85%. Despite improvements in both diagnosis and treatment, the overall prognosis for patients with NSCLC has hardly improved over time; the median survival is 8 months, and the 5-year survival is reported to be <15%.1 Patients with early stages I and II disease who undergo surgical resection have the best chance of survival. To our knowledge, no curative treatment is currently available for patients with stage IV lung cancer. Because the incidence of stage IV NSCLC is rather high, and the expected survival relatively short, this patient category requires a high quality of supportive or palliative care.
Quality of care depends largely on 3 perspectives: professional quality, organizational quality, and patient-oriented quality. An “integrated care” approach is obtained when patients receive the right diagnostic procedures and treatment options according to the best available evidence (professional quality), when the coordination of care is optimal and cooperation between professionals is such that it minimizes duplications and speeds up throughput times (organizational quality), and when patients are optimally involved and supported in their care (patient-oriented quality).2 Actual care in all 3 dimensions should be assessed to determine whether high-quality integrated care is being delivered to patients with NSCLC and to learn whether there is room for improvement.
To our knowledge, good data reflecting the actual care that encompasses the 3 perspectives are currently not available. However, to meet the targets of integrated care, a good understanding of both actual performance and factors determining variations in performance are needed.3 Quality indicators are needed to measure performance. Quality indicators are defined as measurable elements of practice performance for which there is evidence or consensus that they can assess the quality (and therefore a change in quality) of the care provided.4 These indicators should be based on available evidence and include the 3 perspectives of professional, organizational, and patient-oriented quality. Recently, we developed a set of quality indicators for patients with NSCLC based on evidence-based guidelines and the opinions of professionals and patients.5
Determinants of either high or low quality of care for patients with NSCLC can be related to patient, professional, or hospital characteristics, such as patient age,6 health status,7 stage of disease, comorbidity,8 years of experience of the professionals involved,9 their knowledge of guidelines and barriers to the implementation of guidelines,10, 11 the size of the hospital and the patient volume treated,12, 13 teaching activities in the hospital,14 and the availability and functioning of a multidisciplinary lung cancer team or specialized nurses with coordination tasks.2, 15 Insight into determinants that influence the quality of care can help to design targeted strategies with which to improve care. However, to our knowledge, such factors have never been studied with respect to patients with NSCLC.
In the current study, we focused on determinants influencing the quality of care and the variations in the actual quality of integrated care for patients with NSCLC to estimate whether there is room for improvement.
MATERIALS AND METHODS
Setting and Study Population
The quality of integrated care for patients with NSCLC was assessed in 6 hospitals in the eastern region of the Netherlands: 1 university hospital, 4 teaching hospitals, and 1 nonteaching hospital. The mean numbers of patients seen annually who were diagnosed with NSCLC varied from 30 patients in the smallest hospital to 160 in the largest hospital.16 We selected patients with NSCLC (based on the World Health Organization [WHO] classification) using the pathology diagnoses collected by the East Comprehensive Cancer Centre (CCCE). All consecutive newly diagnosed patients with NSCLC were included in the study during a 6-month period from September 2004 to February 2005. All specialists and specialist nurses involved in the care of these patients were also included. The regional research ethics committee approved the study and permission to collect additional data was obtained from the contacts in each hospital.
Quality Indicators and Data Collection
The quality of integrated care for patients with NSCLC was measured with a set of indicators that were systematically developed and validated as described in a previous study.5 This set of indicators was based on national and international evidence-based guidelines for the diagnosis and treatment of patients with NSCLC, and includes the opinions of patients and professionals, and consists of 14 indicators: 4 regarding professional evidence-based quality, 3 regarding organizational quality, and 7 regarding patient-oriented quality (Table 1).
|Quality indicator||Score %||Range %|
|Professional quality (n = 276)|
|1||No. of patients with surgery who underwent mediastinoscopy preceded by FDG-PET (58)||88||83–100|
|Total no. of patients with surgery (66)|
|2†||No. of patients who had a cervical mediastinoscopy performed according to the guideline criteria (219)||84||68–100*|
|Total no. of patients (260)|
|3†||No. of patients who underwent a cervical mediastinoscopy and biopsies of at least 4 of the 6 accessible lymph node stations (24)||50||0–71*|
|Total no. of patients who underwent a cervical mediastinoscopy (48)|
|4||No. of patients with clinical stage III NSCLC for whom skeletal scintigraphy and a CT or MRI of the brain was performed before the initiation of combination therapy (1)||3||0–20*|
|Total no. of patients with clinical stage III NSCLC who received combination therapy (29)|
|Organizational quality (n = 276)|
|5†||No. of patients who had the diagnostic course completed within 21 calendar d from the time of the first visit to the pulmonologist (chest CT scan, bronchoscopy, FDG-PET) (189) (with mediastinoscopy)||79 (67 with mediastoncopy)||71–84 (58–73)|
|Total no. of patients who underwent diagnostic procedures (239)|
|6†||No. of patients who began therapy within 35 calendar d from the time of the first visit to the pulmonologist (80)||51||38–66*|
|Total no. of patients who began therapy (157)|
|7†||No. of patients discussed during multidisciplinary consultation (available in 2 of 6 hospitals) (156)||57||26–91*|
|Total no. of patients (276)|
|Patient-oriented quality (n = 100)|
|8†||No. of patients reporting that attention was paid to physical symptoms: pain, suffocation, nausea, fatigue, weight loss, and insomnia (52)||58||25–78*|
|Total no. of patients (89)|
|9†||No. of patients reporting that they were asked about psychosocial stress factors and psychologic symptoms (28)||34||18–60*|
|Total no. of patients (83)|
|10†||No. of patients reporting that they were asked about psychosocial problems in family and problems related to living conditions (31)||36||23–60*|
|Total no. of patients (85)|
|11||No. of patients who were in need of psychosocial care from trained providers and received it (11)||39||0–100*|
|Total no. of patients in need of psychosocial care from trained providers (28)|
|12||No. of patients reporting that they were consulted adequately (97)||98||96–100|
|Total no. of patients (99)|
|13†||No. of patients reporting that they were informed about the existence of a oncology nurse specializing in lung cancer treatment (49)||53||33–86*|
|Total no.of patients (93)|
|14†||No. of patients reporting that they were informed adequately about all 10 information aspects (18)||19||17–57*|
|Total no. of patients (95)|
The CCCE registration clerks extracted data that were needed to evaluate the selected recommendations. In summary, these data were: patient and disease characteristics, dates of diagnostic procedures and treatment, and whether patient care was discussed with at least 1 other specialist.
For the indicator “carrying out a cervical mediastinoscopy according the guideline criteria,” a thoracic surgeon extracted the data and recorded it on a special data extraction form. We collected the data for the indicators of patient-oriented quality with patient questionnaires. The patient records had been screened to exclude any patients who had died. We sent questionnaires to only 132 of the 276 patients because 144 patients had died at the time this study was conducted. A reminder was sent 2 weeks later.
Determinants and Data Collection
The potentially relevant patient characteristics measured in the current study were age, comorbidity (chronic heart failure, chronic lung disease, diabetes mellitus, and other malignancies present >5 years previously), and disease stage (according to the TNM classification).17 The CCCE routinely extracted these data from the medical records.
We collected the professional characteristics from the individual questionnaires sent to all 79 medical specialists and specialist nurses who were involved in the care of patients with lung cancer. The questionnaire was comprised of 4 parts. The first part included demographic data (age and sex), professional background, and the clinical experience of the specialist. Parts 2 and 3 included questions to assess “knowledge of the NSCLC guideline” and “barriers for guideline implementation.” These questions were based on a validated questionnaire developed at our center18; “knowledge of the NSCLC guideline” was assessed with 9 items regarding attitudes toward the Dutch national NSCLC guideline and 17 items concerning “barriers for guideline implementation.” We assessed team climate among all the professionals involved in the care of patients with NSCLC by means of the short-form Team Climate Inventory (TCI) developed by Anderson and West.19
We obtained the hospital characteristics from the routinely collected CCCE data (number of patients seen annually who were diagnosed with NSCLC) and from a structured interview with the contact person (a pulmonologist) at each hospital. The hospital characteristics consisted of teaching status, availability of a multidisciplinary team with all specialists present as reported in the NSCLC guideline, and the availability of ≥1 specialized nurses with coordination tasks.
We calculated the frequencies to evaluate scores on the 14 integrated care indicators and the variations among the 6 hospitals. Two indicators had high scores and less variation among the hospitals (Indicators 1 and 12, Table 1) and 2 indicators involved <30 cases (Indicators 4 and 11, Table 1). These 4 indicators were excluded from determinant analysis.
We studied the single correlations between adherence to the 10 remaining quality indicators and all possible influential determinants (at the patient, professional, and hospital levels) using univariate analysis (chi-square test and Student t test). The influential factors that were found to be suitable for further analysis (P < .15) were tested on intercorrelations. If a correlation was detected between 2 independent variables (correlation coefficient >0.4), only 1 variable was included in the multilevel analysis. We used multilevel logistic regression analysis to assess which determinants influenced the quality indicators. We constructed separate multivariate backwards stepwise logistic regression models in which each of the quality indicators formed the dependent outcome and all the patient, professional, and hospital characteristics that had bivariate associations (with P < .15) were the independent variables. We composed a random coefficient model with a GLIMMIX procedure using SAS software (SAS for Windows, version 8.2; SAS Institute Inc, Cary, NC). We calculated the percentage of variance that the determinants could explain for each quality indicator. We used a method based on a threshold model to compute the explained variance. Odds ratios (ORs) described the associations between characteristics and quality indicators. An OR > 1 indicates a positive association with the indicator. We considered 2-sided P levels of < .01 to be statistically significant.
A total of 276 patients were included in the current study. We measured the patient-centered indicators from the questionnaires that were returned by 100 of the 132 recipients (response rate of 76%). Of the 79 professionals involved in the care of patients with NSCLC, 55 returned the questionnaire (response rate of 70%).
Table 2 outlines the characteristics of patients, professionals, and hospital settings. The mean age of the patients was 67 years (range 34–91 years), 63% had ≥1 comorbid disorders, and 44% of the patients were classified as having stage IV disease. Of the 276 patients listed, 26% had undergone surgery as their initial treatment. The professional characteristics showed that 40% of the professionals had >10 years of experience as medical specialists, that the majority of specialists were pulmonologists (36%), and that 60% were men. Only 2 of the 6 hospitals had a weekly multidisciplinary lung cancer meeting with all their professionals, which is required according to the guideline. One hospital had specialized nurses with coordinating tasks for patients with lung cancer. The mean score obtained on the TCI was 54, of the maximum 70 points (77%). There was little variation noted among hospital teams with regard to TCI scores (highest and lowest scores were 57 and 48, respectively).
|Patients (n = 276)|
|Mean age, y||67||(34–91)|
|Professionals (n = 55)|
|Mean age, y||45||(32–63)|
|>10 y of experience as a specialist, mean no. of y||22||40 (0.5–31)|
|Hospitals (n = 6)|
|Mean no. of hospital beds, (range)||630||(200–960)|
|No. of lung cancer patients (NSCLC) seen per year, (range)||100||(30–160)|
|Availability of a multidisciplinary team for lung cancer treatment||2||33|
|Specialized nurse with coordination tasks||1||17|
|Team climate, total mean score on the TCI†||54||78 (48–57)|
Table 1 shows the scores on the quality indicators and the variation in scores noted among the 6 hospitals. The highest score in the dimension of professional quality was the proportion of patients who had undergone fluorodeoxyglucose-positron emission tomography (FDG-PET) (88%) followed by the number of patients who had undergone cervical mediastinoscopy according the guideline (84%). The lowest scores were for “the number of patients with clinical stage III NSCLC for which skeletal scintigraphy and [computed tomography] CT or [magnetic resonance imaging] MRI of the brain was done before the start of the combination therapy” (score of 3%; the lowest score was 0% and the highest score was 3%) and “number of patients who had a cervical mediastinoscopy and biopsies of at least 4 of the 6 accessible lymph node stations” (score of 50%; the lowest score was 0% and the highest score was 71%). With regard to the organizational quality, nearly 80% of the study population completed the diagnostic route within 21 calendar days after the first visit to the pulmonologist (score of 79%; the lowest score was 71% and the highest score was 84%; the score was 67% if including mediastinoscopy), half of the patients began therapy within 35 days after the first visit (score of 51%; the lowest score was 38% and the highest score was 66%), and 57% of the patients were discussed during multidisciplinary consultation (the lowest score was 26% and the highest score was 91%).
All patient-oriented quality-of-care indicators regarding physical, emotional, and psychosocial care and information supplied were found to have very low scores (≤58%), with the exception of the indicator “number of patients who reported that they were consulted properly” (98%). There was a variation noted of >20% in 6 of the 7 patient-oriented indicators (Table 1).
Overall, there was a variation in quality scores of ≥20% noted among the hospitals for 11 of the 14 indicators.
Table 3 shows all the determinants that remained after multilevel regression analysis. None of the tested determinants was found to have a significant influence on indicators within the dimension of professional quality. For the other indicators, some were found to have a significant influence on ≥1 quality indicators. The 3 indicators in the area of organizational quality were all found to be significantly associated with 1 or 2 patient characteristics. More patients with a higher stage of disease (stage III or IV) completed the diagnostic route within 21 days (odds ratio [OR], 2.71; 95% confidence interval [95% CI], 1.35–5.44) and began therapy within 35 days (OR, 2.83; 95% CI,1.38–5.82). As expected, the greatest predictor of patients being discussed during multidisciplinary consultation is the availability of a multidisciplinary team for lung cancer treatment (OR, 8.80; 95% CI, 2.20–35.22). Older patients and those with a more advanced stage of disease were less likely to be discussed during multidisciplinary consultation (OR, 0.94; 95% CI, 0.92–0.97 and OR, 0.19; 95% CI, 0.09–0.42, respectively).
|Quality indicator||Predictor||OR (95% CI)*||Significance||Explained variance, %†|
|Cervical mediastinoscopy performed according to the guideline criteria||Greater knowledge of the guidelines||0.34 (0.11–1.09)||.07||7.3|
|Diagnostic trajectory completed within 21 calendar d from the time of the first visit to the pulmonologist (chest CT scan, bronchoscopy, FDG-PET)||Higher stage of disease||2.71 (1.35–5.44)||.005‡||5.6|
|Therapy initiated within 35 calendar d from the first visit to the pulmonologist||Higher stage of disease||2.83 (1.38–5.82)||.005‡||5.9|
|Lung cancer discussed during multidisciplinary consultation||Availability of a multidisciplinary team||8.80 (2.20–35.22)||.002‡||38.9|
|Older age||0.94 (0.92–0.97)||.000‡|
|Higher stage of disease||0.19 (0.09–0.42)||.000‡|
|Attention to physical symptoms||Specialized nurses with coordination tasks||0.19 (0.05–0.78)||.02||9.9|
|Attention to psychosocial stress factors and psychologic symptoms||>100 patients seen each y||0.46 (0.17–1.21)||.11||3.1|
|Attention to psychosocial problems in family and problems related to living conditions||Older age||0.96 (0.92–0.99)||.05||7.1|
|Patient informed regarding existence of nurses specialized in lung cancer treatment||Comorbidity||0.24 (0.09–0.65)||.005‡||18.5|
|More than 100 patients seen each y||0.26 (0.08–0.80)||.02|
|Patient adequately informed regarding 10 information aspects||Barriers to guideline implementation||0.47 (0.15–1.52)||.20||3.7|
With regard to the association between determinants and patient-oriented quality indicators, the only characteristic that was found to achieve statistical significance was “comorbidity of patients.” Patients with a comorbidity were less informed regarding the existence of a oncology nurse specializing in lung cancer care (OR, 0.24; 95% CI, 0.09–0.65).
To our knowledge, the current study is the first to examine the variation and determinants of quality of integrated care for patients with NSCLC that uses 3 dimensions of care: professional quality, organizational quality, and patient-oriented quality. The results of the current study demonstrated that there was a large variation between hospitals with regard to scores for the quality indicators, and that the quality of integrated care (especially patient-oriented care) for NSCLC patients needs improvement. Patient characteristics were proven to have more influence on the scores than professional or hospital characteristics; patients with more advanced disease went through the diagnostic course more quickly and began therapy earlier, but were less likely to be discussed during multidisciplinary consultation. Patients with a comorbidity tended to be less informed regarding the existence of oncology nurses specializing in lung cancer care.
Mediastinoscopy is used for intrathoracic staging of the mediastinum and determining lymph node involvement. The results of the current study demonstrate a high score (84%) for performing (17%) or skipping (67%) mediastinoscopy according to the criteria provided in the guideline. This was higher than the results of another Dutch study, namely, adherence to the guidelines for approximately 67% of the patients.20 A further analysis of the cases that deviated from the guideline criteria demonstrates that the main reason for not performing a mediastinoscopy was the use of alternative methods for staging of the mediastinum (transbronchial needle aspiration or endoscopic ultrasound-guided fine-needle aspiration).21, 22 However, because of their recent appearance, these techniques have not been incorporated into the guidelines on which the indicators were based. Obviously, some physicians have already implemented these new techniques, although the guideline committee has not yet formally accepted them.
The low score for adequate sampling of lymph nodes (50%) was in agreement with the results of another study in the Netherlands that demonstrated that mediastinoscopy was performed according to guideline criteria in 40% of cases.20 In addition, the study shows that the hospital with the least number of patients had the lowest score, and this association did not remain on multilevel analysis. We also expected to find some positive influences on the professional quality scores of determinants such as “years of experience of professionals” and “teaching status of the hospital” because there were large variations noted among the hospitals with regard to adequate sampling scores. Although we did indeed find these correlations on univariate analysis, they did not remain significant on multivariate analysis. The small number of hospitals may play a part.
In the current study, only 29 of the 69 patients with clinical stage III NSCLC received combination therapy, and only 1 patient had undergone both skeletal scintigraphy and CT or MRI of the brain before the initiation of combination therapy. All 6 hospitals had very low scores for this indicator. Of the patients who had not undergone skeletal scintigraphy, 50% had undergone a PET scan, which is quite likely the reason that no subsequent bone scan was performed. Possible other reasons for these low scores are the expected low gain of 5% (not everyone is convinced that this is a cost-effective strategy23, 24) and that waiting times for CT and MRI are usually long, which slows down the throughput times before treatment can be initiated.
In agreement with the study of Salomaa et al., we found that only half of the patients received their initial therapy within 35 days of their first visit.25 The variation among the hospitals of 28% showed possibilities for improving this aspect of care. In the current study, delays were found to be shorter for patients with more advanced cancer. The literature shows that long waiting times do not by definition lead to poorer survival.25 However, several studies reported tumor growth during waiting times that eventually led to patients with curative options becoming incurable.26, 27 From the patient's point of view, long waiting times for diagnosis and treatment are unacceptable. There is some early literature regarding the positive effects of so-called “early diagnosis clinics” on reducing waiting and throughput times.28
Only 2 of the 6 hospitals in the current study had a multidisciplinary team for lung cancer, which is required by guideline criteria. This is the main reason for the large variation noted among the hospitals regarding patients being discussed during multidisciplinary consultations (the lowest score was 26% and the highest score was 91%). As expected, the results of the current study demonstrate that patients with a higher stage of disease (stage III or stage IV) and presumably monodisciplinary treatment options are less likely to be discussed during multidisciplinary consultation. Although to our knowledge there is only limited evidence regarding the effects on patient outcomes of discussing patients as part of a multidisciplinary team,29, 30 the majority of cancer guidelines recommend that all patients with cancer need to be presented in a multidisciplinary team at least once.31–33
The 7 patient-centered indicators were based on 2 recommendations from the guidelines. These recommendations were selected by an expert panel as being most important for the patient-oriented quality of care. The low scores on these indicators were striking. Especially for patients with lung cancer, palliative and supportive care is of great importance because the majority of patients who are diagnosed with lung cancer have incurable disease. In agreement with the literature, patients in the current study reported that, in their view, physical symptoms and psychosocial problems were poorly evaluated and managed.34 The impact of lung cancer and its symptoms on the patient's psychologic, social, and physical state should be identified early, and patients should be referred to the appropriate specialist for further assessment as required. The literature shows that screening lists for quality-of-life issues could be helpful35 and that structured follow-up by nurses can improve psychosocial functioning.36
Patients with greater comorbidity appeared to be less informed regarding the existence of a nurse specializing in lung cancer care. The main reason for this is that only 1 hospital actually had a specialized lung cancer nurse. Another possible explanation could be that patients with comorbidity most likely have appointments with many specialists, which may be a reason to forget details from consultations.
The results of the current study demonstrated a high number of patient and professional responses to our questionnaires. However, the finding that greater than half of the patients in the study group had died and therefore could not be questioned most likely affected the results of the patient-oriented quality indicators. The literature appears contradictory regarding the possible effects of stage of disease on the patients' assessments of care.6, 7, 37
The scores on the indicator regarding mediastinoscopy performed according to guideline criteria is a good example of the finding that a low score on an indicator does not automatically mean that there is a problem in the quality of care, but rather is a signal to further evaluate the matter. There can either be new evidence that demands a change in recommendations, or care should be improved. In the Netherlands, the results of the current study and newly available evidence regarding staging of the mediastinum led to the adaptation of the national guideline. An important message for the medical community is that indicators as well as guidelines should be periodically updated; as such, we call it ‘having a living guideline’ that ideally is being updated continuously as new evidence becomes available. Low scores also can be a signal that care should be improved. The results of the current study demonstrate that agreeing on best clinical practice does not automatically lead to implementation of the recommendations in the field. The best guaranteed quality of care could be achieved only by measuring the quality of care from different dimensions, followed by an implementation program (eg, giving feedback to professionals) and then by measuring the quality of care once again.
We conclude that the quality of integrated care for patients with NSCLC can be improved in all dimensions, but special attention to the patient-oriented issues is needed. With regard to determinants that influence the quality of care, the results of the current study demonstrate that patient characteristics appear to have more influence than professional or hospital characteristics. The next step in improving the quality of care for patients with NSCLC is the development of improvement strategies. These implementation programs should be targeted at specific subgroups. Especially for patients with curative treatment possibilities, diagnostic and therapeutic delays should be minimized and patients with comorbidities in particular should be informed regarding the existence of oncology nurses specializing in lung cancer care.
We thank the contact persons of the 6 participating hospitals for their cooperation: M. van der Drift (Radboud University Hospital, Nijmegen), H. Smit (Rijnstate Hospital, Arnhem), R. Termeer (Canisius Wilhelmina Hospital, Nijmegen), R. Bunnik (Pantein Hospital, Boxmeer), M. Oudijk (Gelders Valley Hospital, Ede), and G. Bosman (Slingeland Hospital, Doetinchem)
We also thank the registration clerks of the East Comprehensive Cancer Centre for collecting the data and R. Akkermans for his statistical analyses.
- 3Improving Patient Care: The Implementation of Change in Clinical Practice. London: Elsevier, Butterworth, Heinemann; 2005., , .
- 4Indicators of quality in health care. Eur J Gen Pract. 1997; 3: 103–108., .
- 9Does clinical experience with a treatment regimen affect survival of lung cancer patients? An analysis based on consecutive randomized trials of the Medical Research Council in small cell and non-small cell tumours.Medical Research Council Lung Cancer Working Party. Clin Oncol (R Coll Radiol). 1994; 6: 81–90., , , , , .
- 16Dutch Comprehensive Care Centres. Available at URL: www.ikcnet.nl. Accessed March 2007.
- 17SobinLH,WittekindCH, editors. TNM Classification of Malignant Tumours.6th ed. New York: John Wiley & Sons; 2002.
- 18Room for improvement? Barriers and opportunities for improving patient care. Nijmegen, The Netherlands: Centre for Quality of Care Research; 2003., , , .
- 31Comprehensive Care Centres. Available at URL: www.onco line.nl. Accessed March 2007.
- 32National Guideline Clearinghouse (NGC). Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services. Available at URL: www.guideline.gov. Accessed March 2007.
- 33National Institute for Health and Clinical Excellence of England. Available at URL: www.nice.org.uk. Accessed March 2007.