The effects of clinical and sociodemographic factors on survival, resource use and lead times in patients with high-grade gliomas: a population-based register study

The effects of clinical and sociodemographic factors on survival, resource use and lead times in... Background Previous studies indicate an effect of sociodemographic factors on risk for being diagnosed with, as well as on survival of cancer in general. Our primary aim was to analyse sociodemographic factors, resource use and lead times in health care after diagnosis with high grade malignant glioma (HGG) in a large population based cohort. Methods A register-based study using several unique high-coverage registries. All patients over the age of 18 diagnosed with HGG in the Swedish Stockholm–Gotland region between 2001 and 2013 (n = 1149) were included. Results In multivariable cox proportional hazard model of survival, older age, male sex and high tumour grade were asso- ciated with worse survival. No significant differences could be seen related to country of birth. A high disposable income was associated with better survival and fewer occasions of pre-diagnostic inpatient care. Older age and comorbidities were correlated with a significantly increased number of outpatient visits the year before HGG diagnosis. In addition, male sex, being born outside Sweden was associated to a higher number of outpatient visits the year after diagnosis in multivariable analysis. Leadtime from diagnosis (first suspicion on brain scan) to surgery showed that the oldest patients, patients with comorbidity and patients born outside Europe had to wait longer for surgery. Conclusions Sociodemographic factors like education, income and country of birth have impact on care processes both before and after the diagnosis HGG. This needs to be acknowledged in addition to important clinical factors like age, comorbidity and tumour grade, in order to accomplish more equal cancer care. Keywords Glioma · Brain neoplasm · Comorbidity · Registries · Health resources · Sociodemographic factors Introduction Diagnosis and treatment of high grade glioma (HGG) have been improved during the last years, still the prognosis is poor. In fact, the 5-year survival for the around 400 patients * Jenny Bergqvist annually diagnosed with HGG in Sweden (total popula- Jenny.bergqvist@ki.se tion 10 million), is lower than 10% [1]. Treatment is based on patient and disease specific prognostic factors like age, Institution of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden performance status, histologic grade and tumour molecular profile. Characteristics of the tumour/disease, health status Capio St Görans Hospital, St Görans plan 1, 112 81 Stockholm, Sweden of the patient as well as treatment regimen clearly affects the outcome [2, 3]. Previous publications suggest that soci- Ivbar Institute AB, Hantverkargatan 8, Stockholm, Sweden 4 odemographic factors are associated with how individuals Medical Management Centre, Karolinska Institutet, respond to, acknowledge symptoms and thereby also affect Stockholm, Sweden 5 time to diagnosis [4]. In addition, an increasing number of Regional Cancer Centre Stockholm Gotland, Stockholm reports present that comorbidities and sociodemographic County Council, Stockholm, Sweden 6 factors such as education, income level or country of birth, Department of Radiation Sciences and Oncology, University of Umeå , Umeå, Sweden Vol.:(0123456789) 1 3 600 Journal of Neuro-Oncology (2018) 139:599–608 not only affects the risk of getting a disease [ 5, 6] but may of birth) [16]. The classification of HGG is based on the also have an impact on survival [7–11]. WHO criteria from 2007 [19]. The regional Ethical Review The risk of treatment bias based on old age for patients Board in Stockholm approved the study protocol (Dnr with glioblastoma as well as for other types of cancer is well 2012/1236-31/4). known, since older cancer patients tend to be offered less aggressive treatments. However, omitting radiotherapy in Study variables the treatment of glioblastoma seem to be associated with, not only older age, but also with race, unmarried status and Health outcomes lower annual income [12]. However, the studies of HGG in relation to sociodemographic factors are relatively few, Survival analysis was performed by calculating the number often based on relatively small patient cohorts and rarely of days from diagnosis until date of death or date of loss-to- include resource use. In addition, they show some contra- follow up based on information from the National Cause of dictory results. Death Register. Therefore, this study was performed on a large high qual- ity unique database combining data from multiple registries Resource use with detailed information (e.g. all diagnosis set and proce- dures performed at hospitals, cause and date of death, soci- Information regarding diagnoses and procedures in inpa- odemographic information and patient and tumour charac- tient and outpatient care, only available for patients living teristics) on all citizens. in Stockholm (n = 845), was extracted from PAS for the Our primary aim was to analyse sociodemographic fac- Stockholm region. Inpatient days and outpatient visits 1 year tors in relation to survival, use of health care resources and before and 1 year after the date of diagnosis were calculated. lead times in the care process for Swedish patients with Inpatient care are days spent in hospital during hospital HGG. admissions. Outpatient care are visits to an outpatient spe- cialist clinic (often to a doctor, but can include visits to other professions like nurse etc.), but not to the primary care or Methods general practitioner. Only patients living in Stockholm at the time of diagnosis were included in analyses of resource use. Study population and data sources Care process: lead times This register-based study included data from national and regional databases with time of recruitment based on the Analysis of lead times (in days) between date of diagnosis time period 2001 to 2013. However, throughout 2015 was to date of surgery as well as between date of surgery to start used to follow patients in terms of resource use and survival of non-surgical cancer treatment (radiotherapy, chemother- after data of diagnosis.The study population was all patients apy or other non-surgical cancer treatments), all reported in the Stockholm region, diagnosed with high-grade glioma in SBTR. according to SNOMED histopathological classification [13] reported in the Swedish Cancer Register (SCR). The SCR Patient characteristics keeps record of all newly detected tumours in Sweden and has a coverage rate above 95% for malignant tumours of A set of clinical and sociodemographic variables was defined which 99% are histologically confirmed [14]. All patients (age, year of diagnosis, sex, comorbidity, educational level, diagnosed between 2001 and 2013 and identie fi d in the SCR income level, country of birth and tumour grade) as relevant [15] were included. The Stockholm region includes around when studying the effect of clinical and sociodemographic 23% of the Swedish population (2.3 million citizens 2015) factors on outcome and resource use. Four pre-defined time [16]. periods were used when stratifying patients by year of diag- Through record-linkage using the patients’ personal nosis (2001–2004, 2005–2007, 2008–2010 and 2011–2013). identification number, data were extracted from: The Swed- Data on age at diagnosis (defined in four categories: 18–39, ish Cancer Register, the National Cause of Death Registry 40–59, 60–69, 70–) and sex were taken from SCR. The (information on date of death) [17], Patient administrative category high-grade glioma includes both grade III (e.g. systems (PAS; information on all healthcare visits and pro- anaplastic astrocytoma) and grade IV gliomas (e.g. glio- cedures in inpatient and outpatient care) [18], the Swed- blastoma multiforme), which are known to have different ish Brain Tumour register (SBTR) (cancer treatment in prognosis. For comorbidity analyses the Elixhauser comor- detail and lead times) [1] and Statistics Sweden’s popula- bidity index was used [20], which consists of a predefined tion data (educational level, disposable income and country set of 31 comorbidity categories and data was extracted from 1 3 Journal of Neuro-Oncology (2018) 139:599–608 601 PAS. Patients were classified as comorbid if at least one The total study population and the Stockholm subpopula- comorbidity diagnosis was registered in the PAS (inpatient, tion had similar characteristics. Average age at diagnosis was outpatient or primary care) during two years before cancer approximately 57 years, range of 19–92 years. The major- diagnosis. Information of educational level (categorized ity of patients, approximately 60% were men. In total, 375 as elementary, high school diploma or university degree), (44%) of the patients suffered from at least one comorbid- disposable income (adjusted for family constellation and ity based on the Elixhauser comorbidity index definition. categorized as low, medium or high income) and country Hypertension was the most common comorbidity followed of birth (categorized as born in Sweden, born in Europe by neurological symptoms/diseases (ataxia, degenerative (not Sweden) or born outside Europe) was obtained from diseases, Parkinson’s disease, MS, epilepsy, tremor), depres- Statistics Sweden. sion, other tumours and diabetes (data not shown). Over 80% (687) were born in Sweden and only 6% (49) were born Statistical analysis outside Europe. A university degree was registered for 328 (44%) of the patients. Further details of the study population Unadjusted survival over time was estimated using in total and the subpopulation in Stockholm are summarised Kaplan–Meier analysis and stratified by age, comorbidity, in Table 1. sex, tumour grade, educational level and disposable income level. Tests of statistical significance of differences between Survival groups were performed using log-rank test for equality of survivor functions. The univariable and multivariable effect Kaplan Meier graphs of survival stratified by age, sex, of a number of selected clinical and sociodemographic vari- comorbidity, education level, income and tumour grade ables (age, year of diagnosis, sex, comorbidity, educational are shown in Fig.  1. In univariable analysis, age, comor- level, income level, country of birth and tumour grade) on bidity, educational level and tumour grade was significantly survival, resource use and lead times was evaluated. A Cox associated with survival (Table 2). Patients 70 years old or proportional hazards regression model was used to calculate older had significantly worse survival (median: 258 days) adjusted hazard ratios (HRs) and 95% CIs for the univari- compared with younger patients (median 1105  days for able as well as multivariable effect on all-cause mortality. age 18–39), p < 0.001. Those with low educational level A negative binomial regression model was used to estimate had worse survival compared with patients with high edu- the effect (univariable and multivariable) of the same set of cational level (median 385 days compared with 501 days), case mix variables on resource use before and after high- p = 0.04. Median survival for patients with comorbidity was grade glioma diagnosis as well as on a set of important lead 343 days compared with 451 days for those without comor- times (days from diagnosis to surgery, days from surgery to bidity p < 0.001 and patients with high tumour grade (IV) histopathological report and days from surgery to start of survived in median 431 days compared with 729 days for non-surgical cancer treatment). Incidence rate ratios (IRR) those with tumour grade 3, p < 0.001. and 95% confidence intervals as well as p-values for each In a multivariable cox proportional hazard model of sur- case mix factor are reported. IRR should be interpreted as vival, older age (HR 4.25 (2.99–6.02) p < 0.001), male ≥70 the relative difference of days or outpatient visits when the sex (HR 1.24 (1.05–1.46) p = 0.01) and high tumour grade explaining factor is changed by one unit. Statistical analysis (IV) (HR 1.57 (1.23–2.02) p < 0.001) were associated with was carried out using STATA 13.1 (Stata Corporation, Col- worse survival, Table 2. High income was associated with a lege Station, TX). To be included in analyses of resource better survival (HR 0.76 (0.60–0.97) p = 0.02), Table 2. high use after diagnosis, the patients have to be alive after 365 days, (n = 454). Resource use Data on health care resource use was not available for Results patients living outside the region, why analysis of comor- bidities and of health care resource use are based on the Patient characteristics Stockholm population only (n = 845). There were 1149 patients diagnosed with HGG in the Stock- Outpatient visits the year before diagnosis holm Gotland Region during 2001–2013 out of which 845 were living in Stockholm at the time of diagnosis. Dur- The multivariable analysis of outpatient visits was adjusted ing the observation period, 1005 patients died. Total time for age, year of diagnosis, sex, comorbidities, educational from diagnosis until death or end of follow-up was 3044 level, income level, country of birth and histopatho- person-years, with a median follow-up time of 457 days. logical tumour grade, Table 3. Older patients (IRR 2.12 1 3 602 Journal of Neuro-Oncology (2018) 139:599–608 Table 1 Descriptive statistics Variable Category No. patients % No. patients total % of study population in total Stockholm n = 1149 and the Stockholm population n = 845 with data on Number of patients 845 1149 health care resource use Age (average) 57.6 56.6 Age category 18–39 90 (845) 10.7 141 (1149) 12.3 40–59 332 (845) 39.3 456 (1149) 39.7 60–69 284 (845) 33.6 379 (1149) 33.0 70– 139 (845) 16.5 173 (1149) 15.1 Year of diagnosis 01/04 223 (845) 26.4 268 (1149) 23.3 05/07 171 (845) 20.2 222 (1149) 19.3 08/10 224 (845) 26.5 337 (1149) 29.3 11/13 227(845) 26.9 322 (1149) 28.0 Sex (male) 513 (845) 60.7 684 (1149) 59.5 Comorbidity (%)** 375 (845) 44.4 Education level Elementary 153 (818) 18.7 210 (1118) 18.8 High school diploma 337 (818) 41.2 472 (1118) 42.2 University degree 328 (818) 40.1 445 (1118) 39.8 Disposable income Low 306 (818) 37.4 430 (1118) 38.5 Intermediate 337 (818) 41.2 478 (1118) 42.8 High 175 (818) 21.4 220 (1118) 19.7 Country of birth Born in Sweden 687 (841) 81.75 977 (1149) 85.0 Born in Europe (not Sweden) 105 (841) 12.5 125 (1149) 10.9 Born outside Europe 49 (841) 5.8 52 (1149) 4.5 Grade IV 725 (845) 85.8 968 (1149) 84.3 All the data included are extracted from the Swedish Cancer Register, except from comorbidity data, which was collected from the Patient administrative systems. For comorbidity analyses the Elixhauser comorbid- ity index was used **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) (1.52–2.96) p < 0.001) and patients with comorbidities Inpatient care the year before diagnosis (IRR 2.41 (2.05–2.84) p < 0.001) had more prediagnostic visits compared with younger patients and those without Age, comorbidity and year of diagnosis were the only comorbidity. Older patients had in median 8 visits com- significant factors in multivariable analysis with regard pared with 4 for the youngest ones. Furthermore, patients to inpatient care days, Table  4. Older patients (3 vs 1, with comorbidities had 8 visits compared with 3 visits for IRR 4.79 (3.07–7.49) p < 0.001), and patients with 70– those without comorbidity. In addition, patients diagnosed comorbidities, (3 vs 1, IRR 1.67 (1.34–2.08) p < 0.001), later in the study period (2011–2013) had statistically had more days of inpatient care the year before diagnosis. significantly fewer outpatient visits the year before HGG Patients diagnosed later in the study period (2011–2013) diagnosis compared with those diagnosed in 2001–2004 had fewer inpatient care days the year before diagnosis (0 (IRR 0.74 (0.60–0.92) p = 0.007). vs 8, IRR 0.14 (0.10–0.19) p < 0.001) than those diagnosed early (2001–2004). Outpatient visits the year after diagnosis Inpatient care the year after diagnosis Those diagnosed the last period in the study (2011–2013) had more outpatient visits 84 versus 59 (IRR 1.53 In multivariable analysis, age was the only significant fac- (1.20–1.95) p < 0.001) the year after diagnosis, Table  3. tor associated to inpatient care the year after diagnosis. In addition, male sex (76 vs 72; IRR 1.29 (1.08–1.53) Older patients (IRR (2.04 (1.52–2.74) p < 0.001) 40−59− p = 0.004) and being born in Europe, not Sweden (84 vs required more days of inpatient care compared with the 75; IRR 1.40 (1.07–1.83) p = 0.01) was associated with a youngest (18–39 years old), 28 versus 15 days in median, higher number of outpatient visits the year after diagnosis Table 4. in multivariable analysis. 1 3 Journal of Neuro-Oncology (2018) 139:599–608 603 Fig. 1 Kaplan–Meier curves for survival according to age, comorbid- between groups were performed using log-rank test for equality of ity status at diagnosis, sex, tumour grade, educational level or income survivor functions (p = 0.000, p = 0.001, p = 0.121, (age) (comorbidity) (sex) level. Estimated survival rate at 1 year and 5 years after diagnosis is p = 0.000, p = 0.095, p = 0.028) (tumour grade) (educational level) (income level) reported in each graph. Tests of statistical significance of differences 1 3 604 Journal of Neuro-Oncology (2018) 139:599–608 Table 2 Cox proportional Cox proportional hazards Survival (univariable) Survival (multivariable) hazards model of survival (univariable and multivariable Haz. ratio (95% CI) p value Haz. ratio (95% CI) p value analyses) of the total study Age category (ref: 18–39) population n = 1149 in the Stockholm–Gotland region  40–59 2.29 (1.83–2.87) < 0.001 2.15 (1.59–2.90) < 0.001  60–69 3.08 (2.45–3.88) < 0.001 3.13 (2.28–4.29) < 0.001  70– 4.68 (3.61–6.07) < 0.001 4.25 (2.99–6.02) < 0.001 Year of diagnosis (ref: 01/04)  05/07 0.98 (0.81–1.19) 0.853 0.91 (0.72–1.14) 0.399  08/10 1.08 (0.91–1.28) 0.402 0.92 (0.74–1.14) 0.450  11/13 0.95 (0.80–1.15) 0.617 0.84 (0.67–1.06) 0.145 Sex  Male 1.12 (0.99–1.27) 0.074 1.24 (1.05–1.46) 0.010 Comorbidity*  Comorbid 1.33 (1.15–1.54) < 0.001 1.13 (0.97–1.32) 0.127 Educational level (ref: elementary)  High school diploma 0.88 (0.74–1.05) 0.166 0.96 (0.78–1.19) 0.739  University degree 0.83 (0.70–0.99) 0.036 0.99 (0.79–1.24) 0.937 Income level (ref: low)  Medium 1.04 (0.91–1.20) 0.563 0.89 (0.74–1.06) 0.191  High 0.97 (0.81–1.16) 0.749 0.76 (0.60–0.97) 0.024 Country of birth (ref: Sweden)  Born in Europe (not Sweden) 1.13 (0.93–1.39) 0.219 1.23 (0.98–1.55) 0.072  Born outside Europe 0.81 (0.59–1.11) 0.192 0.94 (0.65–1.34) 0.722 Grade IV (ref: grade III)  Grade IV 2.00 (1.58–2.53) < 0.001 1.57 (1.23–2.02) < 0.001 Bold values indicate the p value ≤ 0.05 *Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) with patients born in Sweden and patients with grade III Lead times tumours. Lead times from diagnosis to surgery Lead times from surgery to start of non-surgical cancer treatment The lead time analysis showed differences among patients depending on clinical and sociodemographic factors. The multivariable analysis of time from surgery to start of non- Results of both univariable and multivariable analysis surgical cancer treatment decreased over time and was signifi- are reported in Table  5. The multivariable analysis of cantly shorter during the last half of the study period (30 vs time from diagnosis (first suspicion on brain scan) to 60 days, IRR 0.40 (0.26–0.61) p = < 0.001), Table 5. This surgery showed that the oldest patients (17 vs 9 days, 11/13 waiting time was shorter for those 60–69 years old compared IRR 1.64 (1.10–2.45) p = 0.015), patients with comor- 70− with the youngest patients, 18–39 years old (34 vs 39 days, bidity (15 vs 13 days, IRR 1.53 (1.28–1.83) p = < 0.001) IRR 0.68 (0.52–0.87) p = 0.003). In addition, patients with and patients born outside Europe (15 vs 13 days, IRR 2.79 60−69 grade IV tumours came to start non-surgical cancer treatment (1.91–4.09) p = < 0.001) had to wait longer for surgery earlier compared with those with grade III tumours (35 vs after diagnosis. The same analysis showed that patients 44 days in median, IRR 0.70 (0.55–0.89) p = 0.003) as did born in Europe (not Sweden) (13 vs 14 days, IRR 0.70 patients born outside of Europe compared with those born in (0.55–0.91) p = 0.006) and patients with grade IV tumours Sweden (but similar days in median- 36 vs 35 days, IRR 0.70 (13 vs 17 days, IRR 0.48 (0.36–0.63) p = < 0.001) had (0.50–0.97) p = 0.03). surgery sooner in time from first brain scan, compared 1 3 Journal of Neuro-Oncology (2018) 139:599–608 605 Table 3 The effect of clinical and sociodemographic factors on number of outpatient visits (univariable and multivariable regression analysis) analysed for the Stockholm population, n = 845 Negative binomial regression Preoperative visits (uni- Preoperative visits (multi- Postoperative visits (uni- Postoperative visits (multi- variable) variable) variable) variable) IRR p value IRR p value IRR p value IRR p value Age category (ref: 18–39)  40–59 1.53 (1.15–2.05) 0.004 1.40 (1.06–1.84) 0.016 1.08 (0.86–1.36) 0.487 1.03 (0.81–1.31) 0.814  60–69 2.31 (1.72–3.09) < 0.001 1.82 (1.37–2.41) < 0.001 1.27 (1.00-1.62) 0.052 1.17 (0.89–1.52) 0.256  70– 3.12 (2.25–4.32) < 0.001 2.12 (1.52–2.96) < 0.001 1.13 (0.80–1.58) 0.493 1.07 (0.74–1.54) 0.718 Year of diagnosis (ref: 01/04)  05/07 1.04 (0.82–1.33) 0.748 0.93 (0.75–1.17) 0.556 1.28 (1.02–1.61) 0.035 1.28 (1.00-1.63) 0.049  08/10 0.85 (0.68–1.07) 0.165 0.77 (0.62–0.95) 0.015 1.25 (1.00-1.55) 0.051 1.27 (1.00-1.61) 0.046  11/13 0.94 (0.75–1.18) 0.584 0.74 (0.60–0.92) 0.007 1.46 (1.17–1.82) 0.001 1.53 (1.20–1.95) 0.001 Sex  Male 2.67 (2.30–3.10) < 0.001 0.99 (0.84–1.15) 0.86 1.08 (0.92–1.27) 0.361 1.29 (1.08–1.53) 0.004 Comorbidity 0.80 (0.64-1.00) 0.046 2.41 (2.05–2.84) < 0.001 0.99 (0.78–1.25) 0.906 1.02 (0.86–1.20) 0.846 Educational level (ref: elementary)  High school diploma 0.57 (0.46–0.72) < 0.001 0.96 (0.78–1.19) 0.708 0.99 (0.79–1.26) 0.964 1.01 (0.79–1.28) 0.964  University degree 0.76 (0.63–0.92) 0.004 0.83 (0.67–1.04) 0.114 1.00 (0.83–1.20) 0.968 1.06 (0.83–1.36) 0.657 Income level (ref: low)  Medium 0.78 (0.64–0.93) 0.007 0.86 (0.72–1.03) 0.097 1.02 (0.85–1.23) 0.831 0.93 (0.76–1.12) 0.437  High 0.64 (0.51–0.80) < 0.001 0.81 (0.65–1.02) 0.071 1.03 (0.82–1.29) 0.803 0.83 (0.65–1.06) 0.136 Country of birth (ref:Sweden)  Born in Europe (not Sweden) 1.35 (1.06–1.73) 0.015 1.25 (0.99–1.58) 0.065 1.28 (0.98–1.67) 0.069 1.40 (1.07–1.83) 0.014  Born outside Europe 1.02 (0.72–1.43) 0.93 1.08 (0.77–1.53) 0.647 0.95 (0.68–1.34) 0.77 0.93 (0.64–1.34) 0.681 Grade IV (ref: grade III)  Grade IV 0.00 (0.00–0.00) 0.078 1.08 (0.86–1.35) 0.516 1.22 (0.99–1.51) 0.067 1.10 (0.88–1.39) 0.399 Bold values indicate the p value ≤ 0.05 The analysis of preoperative visits (1 year before diagnosis) include all 845 in the subpopulation but the postoperative visits include only the 454 patients alive after 1 year *IRR = incidence rate ratio **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) with a lower income at a higher risk of earlier death [21]. Discussion One possible explanation to why high income is associated to improved survival may be a better performance status to This is, to our knowledge, the largest population-based study begin with, which the observed requirement of fewer pre- addressing the impact of various clinical and sociodemo- diagnostic visits could suggest. The overall mortality and graphic factors on survival and resource use in patients with morbidity in general is well known to be higher in popula- HGG. It is obvious, that sociodemographic status in addi- tions with low versus high socioeconomic status [22]. tion to patient and disease specific factors did affect survival The lack of any correlation between country of birth and and health care resource use in 845 patients diagnosed with survival are in agreement with other studies, which found no HGG during 2001–2013. A high disposable income was difference in survival for patients with astrocytoma accord- associated with a better survival, while older age, male sex ing to racial disparities [7, 23]. and high tumour grade were associated with worse survival. Field et al. published a multivariable analysis from a com- Country of birth did not show any significant correlation prehensive dataset including 542 patients with glioblastoma. with survival. They report age, poor performance status, operation type and Why females and patients with a higher disposable enrolment in clinical trial to be independent predictors for income have a better chance of survival after HGG diagnosis overall survival in multivariable analysis. In contrast to our can only be speculated on. However, our data is in line with data, they did not find that socioeconomic status, including those published by Sherwood et all, who also found patients 1 3 606 Journal of Neuro-Oncology (2018) 139:599–608 Table 4 The effect of clinical and sociodemographic factors on number of inpatient care days (univariable and multivariable regression analysis) for the Stockholm population (n = 845) Negative binomial regression Preoperative days (uni- Preoperative days (multi- Postoperative days (uni- Postoperative days (multi- variable) variable) variable) variable) IRR p value IRR p value IRR p value IRR p value Age category (ref: 18–39)  40–59 1.36 (0.90–2.05) 0.142 1.62 (1.13–2.31) 0.009 1.96 (1.49–2.59) < 0.001 2.04 (1.52–2.74) < 0.001  60–69 1.75 (1.16–2.66) 0.008 1.91 (1.31–2.79) 0.001 2.53 (1.89–3.40) < 0.001 2.40 (1.72–3.35) < 0.001  70– 2.57 (1.61–4.11) < 0.001 4.79 (3.07–7.49) < 0.001 1.65 (1.09–2.48) 0.017 1.64 (1.04–2.59) 0.033 Year of diagnosis (ref: 01/04)  05/07 0.74 (0.55–1.01) 0.06 0.69 (0.51–0.92) 0.011 1.27 (0.95–1.69) 0.101 1.33 (0.99–1.80) 0.06  08/10 0.25 (0.19–0.34) < 0.001 0.18 (0.13–0.24) < 0.001 1.38 (1.05–1.81) 0.022 1.27 (0.96–1.68) 0.089  11/13 0.25 (0.19–0.34) < 0.001 0.14 (0.10–0.19) < 0.001 1.04 (0.79–1.36) 0.795 1.06 (0.80–1.42) 0.674 Sex  Male 0.93 (0.74–1.17) 0.528 0.91 (0.74–1.13) 0.399 0.95 (0.78–1.16) 0.61 0.93 (0.76–1.14) 0.487 Comorbidity 1.72 (1.37–2.16) < 0.001 1.67 (1.34–2.08) < 0.001 1.28 (1.05–1.56) 0.015 1.07 (0.86–1.33) 0.528 Educational level (ref: elementary)  High school diploma 0.71 (0.52–0.98) 0.039 0.78 (0.58–1.05) 0.101 1.02 (0.76–1.36) 0.919 1.04 (0.78–1.40) 0.785  University degree 0.55 (0.40–0.77) < 0.001 0.84 (0.62–1.15) 0.277 1.02 (0.76–1.36) 0.901 1.09 (0.80–1.47) 0.594 Income level (ref: low)  Medium 0.84 (0.65–1.09) 0.192 0.91 (0.72–1.16) 0.451 1.16 (0.93–1.46) 0.181 1.10 (0.87–1.37) 0.433  High 0.38 (0.28–0.51) < 0.001 0.74 (0.54–1.02) 0.068 1.00 (0.77–1.31) 0.993 0.91 (0.68–1.22) 0.535 Country of birth (ref: Sweden)  Born in Europe (not Sweden) 0.97 (0.68–1.38) 0.868 1.13 (0.82–1.54) 0.463 1.00 (0.73–1.38) 0.994 1.11 (0.81–1.53) 0.505  Born outside Europe 0.90 (0.55–1.46) 0.673 1.10 (0.68–1.79) 0.690 0.84 (0.56–1.27) 0.406 0.82 (0.53–1.28) 0.376 Grade IV (ref: grade III)  Grade IV 0.94 (0.68–1.30) 0.705 1.06 (0.78–1.44) 0.709 1.41 (1.09–1.83) 0.009 1.21 (0.92–1.59) 0.172 Bold values indicate the p value ≤ 0.05 The analysis of preoperative days in hospital (1 year before diagnosis) include all 845 in the subpopulation but the postoperative days in hospital include only the 454 patients alive after 1 year *IRR (CI) = incidence rate ratio with 95% confidence intervals **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) income, had impact on survival [10]. However, the study analysis of resource use confirmed this in univariable analy - populations and included variables are not exactly the same. sis. However, in multivariable analysis comorbidities was not Our data is in line with other studies that older patients an independent prognostic factor, which may be explained have worse survival [24]. However, this study does not by its covariation with some of the sociodemographic factors include analysis of specific treatments given and it is pos- included, such as income and education. sible that one explanation to age being highly significant In addition to comorbidity, other clinical and sociodemo- is simply because older patients are more often excluded graphic factors showed significant differences in relation to from more aggressive and possibly more effective treatments resource use, before and after diagnosis. The year before for HGG [25]. Older patients and those with comorbidities HGG diagnosis, patients with a higher income level as well had twice as many prediagnostic visits before they came to as patients with a higher educational level had both fewer diagnosis with HGG. hospital admissions as well as outpatient visits. This was not The importance of comorbidities is well known in the significant at the multivariable level but may be interesting clinic situation, still little is known to what extent it affects to further investigate in the future. Of course, the situation outcome and resource use for HGG patients. Our previously is complex and multifactorial but we wanted to investigate published data, showed comorbidities to be associated with different sociodemographic factors’ possible associations decreased survival and increased resource use in patients with resource use and delayed diagnosis, which may have with primary brain tumours (not only HGG) [8]. The present impact on survival. 1 3 Journal of Neuro-Oncology (2018) 139:599–608 607 Table 5 The effect of clinical and sociodemographic factors on lead times in care process for the 845 patients in the Stockholm region diagnosed with high grade glioma 2001–2013 Negative binomial regression From diagnosis to surgery From diagnosis to surgery From surgery to start of From surgery to start of (univariate) (multivariate) non-surgical cancer treat- non-surgical cancer treat- ment (univariate) ment (multivariate) IRR (CI) p value IRR (CI) p value IRR (CI) p value IRR (CI) p value Age category (ref: 18–39)  40–59 0.59 (0.44–0.81) 0.001 1.13 (0.82–1.55) 0.455 0.42 (0.33–0.52) 0.000 0.67 (0.53–0.86) 0.002  60–69 0.55 (0.40–0.75) 0.000 1.30 (0.94–1.82) 0.117 0.40 (0.32–0.50) 0.000 0.68 (0.52–0.87) 0.003  70– 0.82 (0.57–1.20) 0.306 1.64 (1.10–2.45) 0.015 0.73 (0.52–1.02) 0.063 1.21 (0.84–1.75) 0.306 Year of diagnosis (ref: 01/04)  05/07 1.57 (1.03–2.38) 0.035 0.91 (0.61–1.35) 0.635 0.43 (0.29–0.62) 0.000 0.64 (0.42–0.98) 0.038  08/10 1.40 (0.93–2.10) 0.105 1.36 (0.92–2.01) 0.127 0.26 (0.18–0.37) 0.000 0.42 (0.27–0.63) 0.000  11/13 0.91 (0.61–1.37) 0.651 0.97 (0.65–1.43) 0.870 0.29 (0.20–0.42) 0.000 0.40 (0.26–0.61) 0.000 Sex  Male 0.83 (0.69-1.00) 0.049 0.91 (0.76–1.08) 0.286 1.25 (1.08–1.46) 0.003 1.01 (0.87–1.18) 0.863 Comorbidity* 1.57 (1.31–1.89) 0.000 1.53 (1.28–1.83) 0.000 0.81 (0.70–0.95) 0.010 0.94 (0.81–1.10) 0.460 Educational level (ref: elementary)  High school diploma 1.00 (0.78–1.29) 0.997 1.11 (0.87–1.41) 0.409 1.19 (0.96–1.48) 0.114 1.06 (0.86–1.31) 0.588  University degree 0.74 (0.57–0.95) 0.018 1.02 (0.79–1.31) 0.879 1.32 (1.07–1.64) 0.011 0.98 (0.78–1.22) 0.843 Income level (ref: low)  Medium 1.30 (1.06–1.60) 0.010 1.04 (0.86–1.27) 0.669 0.68 (0.57–0.82) 0.000 0.87 (0.73–1.04) 0.122  High 0.83 (0.66–1.05) 0.116 0.88 (0.69–1.11) 0.269 0.93 (0.77–1.13) 0.467 1.10 (0.90–1.34) 0.355 Country of birth (ref: Sweden)  Born in Europe (not Sweden) 1.08 (0.82–1.41) 0.000 0.70 (0.55–0.91) 0.006 0.80 (0.63–1.02) 0.077 0.94 (0.76–1.16) 0.565  Born outside Europe 3.43 (2.41–4.89) 0.000 2.79 (1.91–4.09) 0.000 0.78 (0.56–1.09) 0.148 0.70 (0.50–0.97) 0.030 Grade IV (ref: grade III)  Grade IV 0.37 (0.29–0.49) 0.000 0.48 (0.36–0.63) 0.000 0.50 (0.40–0.63) 0.077 0.70 (0.55–0.89) 0.003 Bold values indicate the p value ≤ 0.05 *IRR (CI) = incidence rate ratio with 95% confidence intervals **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) Patients born in Europe, but not in Sweden, had signifi- Another limitation of our paper is that treatment regi- cantly more postoperative outpatient visits. One can specu- men is not included in the analyses, as this stratification late if this may be explained by difficulties with language would render too small subgroups for robust statistical and communication as information, written information is analyses. But we do know that treatment affect survival. usually in Swedish, or if it is due to differences in culture, We also lack information about other factors known to but this needs to be further studied. affect survival like performance status, extent of resection Our analysis of lead times in the care process was and molecular subtypes. On the other hand, one of the employed in order to investigate potential differences strengths is that the analysis includes all patients diag- between patients with different clinical and or/sociode- nosed with HGG between January 1st 2001 and December mographic status. Country of birth did affect time from 31st 2013, in one large region. The unique administrative diagnosis to surgery as well as time from surgery to start database, which covers hospital admissions as well as out- of non-surgical cancer treatment. However, the number of patient visits to specialists and visits in primary care, was patients born in other countries than Sweden, especially out- linked to several registries by the personal identification side Europe, is relatively limited which make it difficult to number. make firm conclusions. And of course, differences in a few Over time we found less need for outpatient visits the days may not be of clinical importance and further studies year before HGG diagnosis but more outpatient visits the are needed to see if the number and extent of differences year after, which can be explained by improved pre diagnos- stays the same. tic care and more interventions after diagnosis. In addition, time from surgery to start of non-surgical cancer treatment 1 3 608 Journal of Neuro-Oncology (2018) 139:599–608 outcomes among elderly individuals with primary malignant decreased over time and was significantly shorter during the astrocytoma. J Neurosurg 108(4):642–648 last half of the study period. 8. Bergqvist J, Iderberg H, Mesterton J, Bengtsson N, Wettermark B, The effect of sociodemographic factors on survival could Henriksson R (2017) Healthcare resource use, comorbidity, treat- potentially be reduced with increased awareness of these ment and clinical outcomes for patients with primary intracranial tumors: a Swedish population-based register study. Acta Oncol inequalities and relocation of resources in order to balance 56(3):405–414 inequities when necessary. For example, we need to increase 9. Danzig MR, Weinberg AC, Ghandour RA, Kotamarti S, McKi- knowledge of when to address the health care system among ernan JM, Badani KK (2014) The association between socioeco- all people regardless of sociodemographic status, like sug- nomic status, renal cancer presentation, and survival in the United States: a survival, epidemiology, and end results analysis. Urology gested by Whitaker et al. [4]. 84(3):583–589 In conclusion, sociodemographic factors have impact on 10. Field KM, Drummond KJ, Yilmaz M, Tacey M, Compston D, survival and resource use for patients with HGG and with Gibbs P et al (2013) Clinical trial participation and outcome for increased awareness and further studies we can reduce the patients with glioblastoma: multivariate analysis from a compre- hensive dataset. J Clin Neurosci 20(6):783–789 inequalities and further reduce clinically relevant differences. 11. Hastert TA, Beresford SA, Sheppard L, White E (2015) Dispari- ties in cancer incidence and mortality by area-level socioeconomic Acknowledgements The authors would like to thank Nils Bengtsson status: a multilevel analysis. J Epidemiol Community Health at Ivbar Institute for work with the database. 69(2):168–176 12. Aizer AA, Ancukiewicz M, Nguyen PL, Shih HA, Loeffler JS, Compliance with ethical standards Oh KS (2014) Underutilization of radiation therapy in patients with glioblastoma: predictive factors and outcomes. Cancer 120(2):238–243 Conflict of interest None to declare for Jenny Bergqvist and Roger 13. Fritz AG (2013) International classification of diseases for oncol- Henriksson. Hanna Iderberg and Johan Mesterton are employed at ogy: ICD-O, 3rd edn, 1st rev. edn. World Health Organization, Ivbar Institute, a research company specialized in health care govern- Geneva, p. viii, 242 ance and analysis of health care data. Johan Mesterton is a shareholder 14. Barlow L, Westergren K, Holmberg L, Talback M (2009) The at Ivbar Institute. completeness of the Swedish Cancer Register: a sample survey for year 1998. Acta Oncol 48(1):27–33 Open Access This article is distributed under the terms of the Crea- 15. National Board of Health and Welfare (2013) The Swedish Cancer tive Commons Attribution 4.0 International License (http://creat iveco Registry (database) mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 16. Statistics Sweden: longitudinal integration database for health tion, and reproduction in any medium, provided you give appropriate insurance and labour market studies (LISA by Swedish acronym) credit to the original author(s) and the source, provide a link to the (2016) Creative Commons license, and indicate if changes were made. 17. The National Board of Health and Welfare (2016) Causes of Death (database) 18. Carlsson AC, Wandell P, Osby U, Zarrinkoub R, Wettermark B, Ljunggren G (2013) High prevalence of diagnosis of diabetes, References depression, anxiety, hypertension, asthma and COPD in the total population of Stockholm, Sweden—a challenge for public health. 1. Asklund T, Malmstrom A, Bergqvist M, Bjor O, Henriksson R BMC Public Health 13:670 (2015) Brain tumors in Sweden: data from a population-based 19. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, registry 1999–2012. Acta Oncol 54(3):377–384 Jouvet A et al (2007) The 2007 WHO classification of tumours of 2. Bauchet L, Mathieu-Daude H, Fabbro-Peray P, Rigau V, Fabbro the central nervous system. Acta Neuropathol 114(2):97–109 M, Chinot O et al (2010) Oncological patterns of care and out- 20. Elixhauser A, Steiner C, Harris DR, Coffey RM (1998) Comor - come for 952 patients with newly diagnosed glioblastoma in 2004. bidity measures for use with administrative data. Med Care Neuro Oncol 12(7):725–735 36(1):8–27 3. Lacroix M, Abi-Said D, Fourney DR, Gokaslan ZL, Shi W, 21. Sherwood PR, Dahman BA, Donovan HS, Mintz A, Given CW, DeMonte F et al (2001) A multivariate analysis of 416 patients Bradley CJ (2011) Treatment disparities following the diagnosis with glioblastoma multiforme: prognosis, extent of resection, and of an astrocytoma. J Neurooncol 101(1):67–74 survival. J Neurosurg 95(2):190–198 22. Chetty R, Stepner M, Abraham S, Lin S, Scuderi B, Turner N et al 4. Whitaker KL, Scott SE, Wardle J (2015) Applying symptom (2016) The Association between income and life expectancy in appraisal models to understand sociodemographic differences in the United States, 2001–2014. JAMA 315(16):1750–1766 responses to possible cancer symptoms: a research agenda. Br J 23. Robertson JT, Gunter BC, Somes GW (2002) Racial differences Cancer 112(Suppl 1):S27-34 in the incidence of gliomas: a retrospective study from Memphis, 5. Chakrabarti I, Cockburn M, Cozen W, Wang YP, Preston- Tennessee. Br J Neurosurg 16(6):562–566 Martin S (2005) A population-based description of glioblas- 24. Thumma SR, Fairbanks RK, Lamoreaux WT, Mackay AR, Dema- toma multiforme in Los Angeles County, 1974–1999. Cancer kas JJ, Cooke BS et al (2012) Effect of pretreatment clinical fac- 104(12):2798–2806 tors on overall survival in glioblastoma multiforme: a surveillance 6. Khanolkar AR, Ljung R, Talbäck M, Brooke HL, Carlsson S, epidemiology and end results (SEER) population analysis. World Mathiesen T, Feychting M (2016) Socioeconomic position and the J Surg Oncol 10:75 risk of brain tumour: a Swedish national population-based cohort 25. Tanaka S, Meyer FB, Buckner JC, Uhm JH, Yan ES, Par- study. J Epidemiol Community Health ney IF (2013) Presentation, management, and outcome of 7. Barnholtz-Sloan JS, Williams VL, Maldonado JL, Shahani D, newly diagnosed glioblastoma in elderly patients. J Neurosurg Stockwell HG, Chamberlain M et al (2008) Patterns of care and 118(4):786–798 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Neuro-Oncology Springer Journals

The effects of clinical and sociodemographic factors on survival, resource use and lead times in patients with high-grade gliomas: a population-based register study

Free
10 pages

Loading next page...
 
/lp/springer_journal/the-effects-of-clinical-and-sociodemographic-factors-on-survival-KwqgbILYCw
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Oncology; Neurology
ISSN
0167-594X
eISSN
1573-7373
D.O.I.
10.1007/s11060-018-2899-0
Publisher site
See Article on Publisher Site

Abstract

Background Previous studies indicate an effect of sociodemographic factors on risk for being diagnosed with, as well as on survival of cancer in general. Our primary aim was to analyse sociodemographic factors, resource use and lead times in health care after diagnosis with high grade malignant glioma (HGG) in a large population based cohort. Methods A register-based study using several unique high-coverage registries. All patients over the age of 18 diagnosed with HGG in the Swedish Stockholm–Gotland region between 2001 and 2013 (n = 1149) were included. Results In multivariable cox proportional hazard model of survival, older age, male sex and high tumour grade were asso- ciated with worse survival. No significant differences could be seen related to country of birth. A high disposable income was associated with better survival and fewer occasions of pre-diagnostic inpatient care. Older age and comorbidities were correlated with a significantly increased number of outpatient visits the year before HGG diagnosis. In addition, male sex, being born outside Sweden was associated to a higher number of outpatient visits the year after diagnosis in multivariable analysis. Leadtime from diagnosis (first suspicion on brain scan) to surgery showed that the oldest patients, patients with comorbidity and patients born outside Europe had to wait longer for surgery. Conclusions Sociodemographic factors like education, income and country of birth have impact on care processes both before and after the diagnosis HGG. This needs to be acknowledged in addition to important clinical factors like age, comorbidity and tumour grade, in order to accomplish more equal cancer care. Keywords Glioma · Brain neoplasm · Comorbidity · Registries · Health resources · Sociodemographic factors Introduction Diagnosis and treatment of high grade glioma (HGG) have been improved during the last years, still the prognosis is poor. In fact, the 5-year survival for the around 400 patients * Jenny Bergqvist annually diagnosed with HGG in Sweden (total popula- Jenny.bergqvist@ki.se tion 10 million), is lower than 10% [1]. Treatment is based on patient and disease specific prognostic factors like age, Institution of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden performance status, histologic grade and tumour molecular profile. Characteristics of the tumour/disease, health status Capio St Görans Hospital, St Görans plan 1, 112 81 Stockholm, Sweden of the patient as well as treatment regimen clearly affects the outcome [2, 3]. Previous publications suggest that soci- Ivbar Institute AB, Hantverkargatan 8, Stockholm, Sweden 4 odemographic factors are associated with how individuals Medical Management Centre, Karolinska Institutet, respond to, acknowledge symptoms and thereby also affect Stockholm, Sweden 5 time to diagnosis [4]. In addition, an increasing number of Regional Cancer Centre Stockholm Gotland, Stockholm reports present that comorbidities and sociodemographic County Council, Stockholm, Sweden 6 factors such as education, income level or country of birth, Department of Radiation Sciences and Oncology, University of Umeå , Umeå, Sweden Vol.:(0123456789) 1 3 600 Journal of Neuro-Oncology (2018) 139:599–608 not only affects the risk of getting a disease [ 5, 6] but may of birth) [16]. The classification of HGG is based on the also have an impact on survival [7–11]. WHO criteria from 2007 [19]. The regional Ethical Review The risk of treatment bias based on old age for patients Board in Stockholm approved the study protocol (Dnr with glioblastoma as well as for other types of cancer is well 2012/1236-31/4). known, since older cancer patients tend to be offered less aggressive treatments. However, omitting radiotherapy in Study variables the treatment of glioblastoma seem to be associated with, not only older age, but also with race, unmarried status and Health outcomes lower annual income [12]. However, the studies of HGG in relation to sociodemographic factors are relatively few, Survival analysis was performed by calculating the number often based on relatively small patient cohorts and rarely of days from diagnosis until date of death or date of loss-to- include resource use. In addition, they show some contra- follow up based on information from the National Cause of dictory results. Death Register. Therefore, this study was performed on a large high qual- ity unique database combining data from multiple registries Resource use with detailed information (e.g. all diagnosis set and proce- dures performed at hospitals, cause and date of death, soci- Information regarding diagnoses and procedures in inpa- odemographic information and patient and tumour charac- tient and outpatient care, only available for patients living teristics) on all citizens. in Stockholm (n = 845), was extracted from PAS for the Our primary aim was to analyse sociodemographic fac- Stockholm region. Inpatient days and outpatient visits 1 year tors in relation to survival, use of health care resources and before and 1 year after the date of diagnosis were calculated. lead times in the care process for Swedish patients with Inpatient care are days spent in hospital during hospital HGG. admissions. Outpatient care are visits to an outpatient spe- cialist clinic (often to a doctor, but can include visits to other professions like nurse etc.), but not to the primary care or Methods general practitioner. Only patients living in Stockholm at the time of diagnosis were included in analyses of resource use. Study population and data sources Care process: lead times This register-based study included data from national and regional databases with time of recruitment based on the Analysis of lead times (in days) between date of diagnosis time period 2001 to 2013. However, throughout 2015 was to date of surgery as well as between date of surgery to start used to follow patients in terms of resource use and survival of non-surgical cancer treatment (radiotherapy, chemother- after data of diagnosis.The study population was all patients apy or other non-surgical cancer treatments), all reported in the Stockholm region, diagnosed with high-grade glioma in SBTR. according to SNOMED histopathological classification [13] reported in the Swedish Cancer Register (SCR). The SCR Patient characteristics keeps record of all newly detected tumours in Sweden and has a coverage rate above 95% for malignant tumours of A set of clinical and sociodemographic variables was defined which 99% are histologically confirmed [14]. All patients (age, year of diagnosis, sex, comorbidity, educational level, diagnosed between 2001 and 2013 and identie fi d in the SCR income level, country of birth and tumour grade) as relevant [15] were included. The Stockholm region includes around when studying the effect of clinical and sociodemographic 23% of the Swedish population (2.3 million citizens 2015) factors on outcome and resource use. Four pre-defined time [16]. periods were used when stratifying patients by year of diag- Through record-linkage using the patients’ personal nosis (2001–2004, 2005–2007, 2008–2010 and 2011–2013). identification number, data were extracted from: The Swed- Data on age at diagnosis (defined in four categories: 18–39, ish Cancer Register, the National Cause of Death Registry 40–59, 60–69, 70–) and sex were taken from SCR. The (information on date of death) [17], Patient administrative category high-grade glioma includes both grade III (e.g. systems (PAS; information on all healthcare visits and pro- anaplastic astrocytoma) and grade IV gliomas (e.g. glio- cedures in inpatient and outpatient care) [18], the Swed- blastoma multiforme), which are known to have different ish Brain Tumour register (SBTR) (cancer treatment in prognosis. For comorbidity analyses the Elixhauser comor- detail and lead times) [1] and Statistics Sweden’s popula- bidity index was used [20], which consists of a predefined tion data (educational level, disposable income and country set of 31 comorbidity categories and data was extracted from 1 3 Journal of Neuro-Oncology (2018) 139:599–608 601 PAS. Patients were classified as comorbid if at least one The total study population and the Stockholm subpopula- comorbidity diagnosis was registered in the PAS (inpatient, tion had similar characteristics. Average age at diagnosis was outpatient or primary care) during two years before cancer approximately 57 years, range of 19–92 years. The major- diagnosis. Information of educational level (categorized ity of patients, approximately 60% were men. In total, 375 as elementary, high school diploma or university degree), (44%) of the patients suffered from at least one comorbid- disposable income (adjusted for family constellation and ity based on the Elixhauser comorbidity index definition. categorized as low, medium or high income) and country Hypertension was the most common comorbidity followed of birth (categorized as born in Sweden, born in Europe by neurological symptoms/diseases (ataxia, degenerative (not Sweden) or born outside Europe) was obtained from diseases, Parkinson’s disease, MS, epilepsy, tremor), depres- Statistics Sweden. sion, other tumours and diabetes (data not shown). Over 80% (687) were born in Sweden and only 6% (49) were born Statistical analysis outside Europe. A university degree was registered for 328 (44%) of the patients. Further details of the study population Unadjusted survival over time was estimated using in total and the subpopulation in Stockholm are summarised Kaplan–Meier analysis and stratified by age, comorbidity, in Table 1. sex, tumour grade, educational level and disposable income level. Tests of statistical significance of differences between Survival groups were performed using log-rank test for equality of survivor functions. The univariable and multivariable effect Kaplan Meier graphs of survival stratified by age, sex, of a number of selected clinical and sociodemographic vari- comorbidity, education level, income and tumour grade ables (age, year of diagnosis, sex, comorbidity, educational are shown in Fig.  1. In univariable analysis, age, comor- level, income level, country of birth and tumour grade) on bidity, educational level and tumour grade was significantly survival, resource use and lead times was evaluated. A Cox associated with survival (Table 2). Patients 70 years old or proportional hazards regression model was used to calculate older had significantly worse survival (median: 258 days) adjusted hazard ratios (HRs) and 95% CIs for the univari- compared with younger patients (median 1105  days for able as well as multivariable effect on all-cause mortality. age 18–39), p < 0.001. Those with low educational level A negative binomial regression model was used to estimate had worse survival compared with patients with high edu- the effect (univariable and multivariable) of the same set of cational level (median 385 days compared with 501 days), case mix variables on resource use before and after high- p = 0.04. Median survival for patients with comorbidity was grade glioma diagnosis as well as on a set of important lead 343 days compared with 451 days for those without comor- times (days from diagnosis to surgery, days from surgery to bidity p < 0.001 and patients with high tumour grade (IV) histopathological report and days from surgery to start of survived in median 431 days compared with 729 days for non-surgical cancer treatment). Incidence rate ratios (IRR) those with tumour grade 3, p < 0.001. and 95% confidence intervals as well as p-values for each In a multivariable cox proportional hazard model of sur- case mix factor are reported. IRR should be interpreted as vival, older age (HR 4.25 (2.99–6.02) p < 0.001), male ≥70 the relative difference of days or outpatient visits when the sex (HR 1.24 (1.05–1.46) p = 0.01) and high tumour grade explaining factor is changed by one unit. Statistical analysis (IV) (HR 1.57 (1.23–2.02) p < 0.001) were associated with was carried out using STATA 13.1 (Stata Corporation, Col- worse survival, Table 2. High income was associated with a lege Station, TX). To be included in analyses of resource better survival (HR 0.76 (0.60–0.97) p = 0.02), Table 2. high use after diagnosis, the patients have to be alive after 365 days, (n = 454). Resource use Data on health care resource use was not available for Results patients living outside the region, why analysis of comor- bidities and of health care resource use are based on the Patient characteristics Stockholm population only (n = 845). There were 1149 patients diagnosed with HGG in the Stock- Outpatient visits the year before diagnosis holm Gotland Region during 2001–2013 out of which 845 were living in Stockholm at the time of diagnosis. Dur- The multivariable analysis of outpatient visits was adjusted ing the observation period, 1005 patients died. Total time for age, year of diagnosis, sex, comorbidities, educational from diagnosis until death or end of follow-up was 3044 level, income level, country of birth and histopatho- person-years, with a median follow-up time of 457 days. logical tumour grade, Table 3. Older patients (IRR 2.12 1 3 602 Journal of Neuro-Oncology (2018) 139:599–608 Table 1 Descriptive statistics Variable Category No. patients % No. patients total % of study population in total Stockholm n = 1149 and the Stockholm population n = 845 with data on Number of patients 845 1149 health care resource use Age (average) 57.6 56.6 Age category 18–39 90 (845) 10.7 141 (1149) 12.3 40–59 332 (845) 39.3 456 (1149) 39.7 60–69 284 (845) 33.6 379 (1149) 33.0 70– 139 (845) 16.5 173 (1149) 15.1 Year of diagnosis 01/04 223 (845) 26.4 268 (1149) 23.3 05/07 171 (845) 20.2 222 (1149) 19.3 08/10 224 (845) 26.5 337 (1149) 29.3 11/13 227(845) 26.9 322 (1149) 28.0 Sex (male) 513 (845) 60.7 684 (1149) 59.5 Comorbidity (%)** 375 (845) 44.4 Education level Elementary 153 (818) 18.7 210 (1118) 18.8 High school diploma 337 (818) 41.2 472 (1118) 42.2 University degree 328 (818) 40.1 445 (1118) 39.8 Disposable income Low 306 (818) 37.4 430 (1118) 38.5 Intermediate 337 (818) 41.2 478 (1118) 42.8 High 175 (818) 21.4 220 (1118) 19.7 Country of birth Born in Sweden 687 (841) 81.75 977 (1149) 85.0 Born in Europe (not Sweden) 105 (841) 12.5 125 (1149) 10.9 Born outside Europe 49 (841) 5.8 52 (1149) 4.5 Grade IV 725 (845) 85.8 968 (1149) 84.3 All the data included are extracted from the Swedish Cancer Register, except from comorbidity data, which was collected from the Patient administrative systems. For comorbidity analyses the Elixhauser comorbid- ity index was used **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) (1.52–2.96) p < 0.001) and patients with comorbidities Inpatient care the year before diagnosis (IRR 2.41 (2.05–2.84) p < 0.001) had more prediagnostic visits compared with younger patients and those without Age, comorbidity and year of diagnosis were the only comorbidity. Older patients had in median 8 visits com- significant factors in multivariable analysis with regard pared with 4 for the youngest ones. Furthermore, patients to inpatient care days, Table  4. Older patients (3 vs 1, with comorbidities had 8 visits compared with 3 visits for IRR 4.79 (3.07–7.49) p < 0.001), and patients with 70– those without comorbidity. In addition, patients diagnosed comorbidities, (3 vs 1, IRR 1.67 (1.34–2.08) p < 0.001), later in the study period (2011–2013) had statistically had more days of inpatient care the year before diagnosis. significantly fewer outpatient visits the year before HGG Patients diagnosed later in the study period (2011–2013) diagnosis compared with those diagnosed in 2001–2004 had fewer inpatient care days the year before diagnosis (0 (IRR 0.74 (0.60–0.92) p = 0.007). vs 8, IRR 0.14 (0.10–0.19) p < 0.001) than those diagnosed early (2001–2004). Outpatient visits the year after diagnosis Inpatient care the year after diagnosis Those diagnosed the last period in the study (2011–2013) had more outpatient visits 84 versus 59 (IRR 1.53 In multivariable analysis, age was the only significant fac- (1.20–1.95) p < 0.001) the year after diagnosis, Table  3. tor associated to inpatient care the year after diagnosis. In addition, male sex (76 vs 72; IRR 1.29 (1.08–1.53) Older patients (IRR (2.04 (1.52–2.74) p < 0.001) 40−59− p = 0.004) and being born in Europe, not Sweden (84 vs required more days of inpatient care compared with the 75; IRR 1.40 (1.07–1.83) p = 0.01) was associated with a youngest (18–39 years old), 28 versus 15 days in median, higher number of outpatient visits the year after diagnosis Table 4. in multivariable analysis. 1 3 Journal of Neuro-Oncology (2018) 139:599–608 603 Fig. 1 Kaplan–Meier curves for survival according to age, comorbid- between groups were performed using log-rank test for equality of ity status at diagnosis, sex, tumour grade, educational level or income survivor functions (p = 0.000, p = 0.001, p = 0.121, (age) (comorbidity) (sex) level. Estimated survival rate at 1 year and 5 years after diagnosis is p = 0.000, p = 0.095, p = 0.028) (tumour grade) (educational level) (income level) reported in each graph. Tests of statistical significance of differences 1 3 604 Journal of Neuro-Oncology (2018) 139:599–608 Table 2 Cox proportional Cox proportional hazards Survival (univariable) Survival (multivariable) hazards model of survival (univariable and multivariable Haz. ratio (95% CI) p value Haz. ratio (95% CI) p value analyses) of the total study Age category (ref: 18–39) population n = 1149 in the Stockholm–Gotland region  40–59 2.29 (1.83–2.87) < 0.001 2.15 (1.59–2.90) < 0.001  60–69 3.08 (2.45–3.88) < 0.001 3.13 (2.28–4.29) < 0.001  70– 4.68 (3.61–6.07) < 0.001 4.25 (2.99–6.02) < 0.001 Year of diagnosis (ref: 01/04)  05/07 0.98 (0.81–1.19) 0.853 0.91 (0.72–1.14) 0.399  08/10 1.08 (0.91–1.28) 0.402 0.92 (0.74–1.14) 0.450  11/13 0.95 (0.80–1.15) 0.617 0.84 (0.67–1.06) 0.145 Sex  Male 1.12 (0.99–1.27) 0.074 1.24 (1.05–1.46) 0.010 Comorbidity*  Comorbid 1.33 (1.15–1.54) < 0.001 1.13 (0.97–1.32) 0.127 Educational level (ref: elementary)  High school diploma 0.88 (0.74–1.05) 0.166 0.96 (0.78–1.19) 0.739  University degree 0.83 (0.70–0.99) 0.036 0.99 (0.79–1.24) 0.937 Income level (ref: low)  Medium 1.04 (0.91–1.20) 0.563 0.89 (0.74–1.06) 0.191  High 0.97 (0.81–1.16) 0.749 0.76 (0.60–0.97) 0.024 Country of birth (ref: Sweden)  Born in Europe (not Sweden) 1.13 (0.93–1.39) 0.219 1.23 (0.98–1.55) 0.072  Born outside Europe 0.81 (0.59–1.11) 0.192 0.94 (0.65–1.34) 0.722 Grade IV (ref: grade III)  Grade IV 2.00 (1.58–2.53) < 0.001 1.57 (1.23–2.02) < 0.001 Bold values indicate the p value ≤ 0.05 *Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) with patients born in Sweden and patients with grade III Lead times tumours. Lead times from diagnosis to surgery Lead times from surgery to start of non-surgical cancer treatment The lead time analysis showed differences among patients depending on clinical and sociodemographic factors. The multivariable analysis of time from surgery to start of non- Results of both univariable and multivariable analysis surgical cancer treatment decreased over time and was signifi- are reported in Table  5. The multivariable analysis of cantly shorter during the last half of the study period (30 vs time from diagnosis (first suspicion on brain scan) to 60 days, IRR 0.40 (0.26–0.61) p = < 0.001), Table 5. This surgery showed that the oldest patients (17 vs 9 days, 11/13 waiting time was shorter for those 60–69 years old compared IRR 1.64 (1.10–2.45) p = 0.015), patients with comor- 70− with the youngest patients, 18–39 years old (34 vs 39 days, bidity (15 vs 13 days, IRR 1.53 (1.28–1.83) p = < 0.001) IRR 0.68 (0.52–0.87) p = 0.003). In addition, patients with and patients born outside Europe (15 vs 13 days, IRR 2.79 60−69 grade IV tumours came to start non-surgical cancer treatment (1.91–4.09) p = < 0.001) had to wait longer for surgery earlier compared with those with grade III tumours (35 vs after diagnosis. The same analysis showed that patients 44 days in median, IRR 0.70 (0.55–0.89) p = 0.003) as did born in Europe (not Sweden) (13 vs 14 days, IRR 0.70 patients born outside of Europe compared with those born in (0.55–0.91) p = 0.006) and patients with grade IV tumours Sweden (but similar days in median- 36 vs 35 days, IRR 0.70 (13 vs 17 days, IRR 0.48 (0.36–0.63) p = < 0.001) had (0.50–0.97) p = 0.03). surgery sooner in time from first brain scan, compared 1 3 Journal of Neuro-Oncology (2018) 139:599–608 605 Table 3 The effect of clinical and sociodemographic factors on number of outpatient visits (univariable and multivariable regression analysis) analysed for the Stockholm population, n = 845 Negative binomial regression Preoperative visits (uni- Preoperative visits (multi- Postoperative visits (uni- Postoperative visits (multi- variable) variable) variable) variable) IRR p value IRR p value IRR p value IRR p value Age category (ref: 18–39)  40–59 1.53 (1.15–2.05) 0.004 1.40 (1.06–1.84) 0.016 1.08 (0.86–1.36) 0.487 1.03 (0.81–1.31) 0.814  60–69 2.31 (1.72–3.09) < 0.001 1.82 (1.37–2.41) < 0.001 1.27 (1.00-1.62) 0.052 1.17 (0.89–1.52) 0.256  70– 3.12 (2.25–4.32) < 0.001 2.12 (1.52–2.96) < 0.001 1.13 (0.80–1.58) 0.493 1.07 (0.74–1.54) 0.718 Year of diagnosis (ref: 01/04)  05/07 1.04 (0.82–1.33) 0.748 0.93 (0.75–1.17) 0.556 1.28 (1.02–1.61) 0.035 1.28 (1.00-1.63) 0.049  08/10 0.85 (0.68–1.07) 0.165 0.77 (0.62–0.95) 0.015 1.25 (1.00-1.55) 0.051 1.27 (1.00-1.61) 0.046  11/13 0.94 (0.75–1.18) 0.584 0.74 (0.60–0.92) 0.007 1.46 (1.17–1.82) 0.001 1.53 (1.20–1.95) 0.001 Sex  Male 2.67 (2.30–3.10) < 0.001 0.99 (0.84–1.15) 0.86 1.08 (0.92–1.27) 0.361 1.29 (1.08–1.53) 0.004 Comorbidity 0.80 (0.64-1.00) 0.046 2.41 (2.05–2.84) < 0.001 0.99 (0.78–1.25) 0.906 1.02 (0.86–1.20) 0.846 Educational level (ref: elementary)  High school diploma 0.57 (0.46–0.72) < 0.001 0.96 (0.78–1.19) 0.708 0.99 (0.79–1.26) 0.964 1.01 (0.79–1.28) 0.964  University degree 0.76 (0.63–0.92) 0.004 0.83 (0.67–1.04) 0.114 1.00 (0.83–1.20) 0.968 1.06 (0.83–1.36) 0.657 Income level (ref: low)  Medium 0.78 (0.64–0.93) 0.007 0.86 (0.72–1.03) 0.097 1.02 (0.85–1.23) 0.831 0.93 (0.76–1.12) 0.437  High 0.64 (0.51–0.80) < 0.001 0.81 (0.65–1.02) 0.071 1.03 (0.82–1.29) 0.803 0.83 (0.65–1.06) 0.136 Country of birth (ref:Sweden)  Born in Europe (not Sweden) 1.35 (1.06–1.73) 0.015 1.25 (0.99–1.58) 0.065 1.28 (0.98–1.67) 0.069 1.40 (1.07–1.83) 0.014  Born outside Europe 1.02 (0.72–1.43) 0.93 1.08 (0.77–1.53) 0.647 0.95 (0.68–1.34) 0.77 0.93 (0.64–1.34) 0.681 Grade IV (ref: grade III)  Grade IV 0.00 (0.00–0.00) 0.078 1.08 (0.86–1.35) 0.516 1.22 (0.99–1.51) 0.067 1.10 (0.88–1.39) 0.399 Bold values indicate the p value ≤ 0.05 The analysis of preoperative visits (1 year before diagnosis) include all 845 in the subpopulation but the postoperative visits include only the 454 patients alive after 1 year *IRR = incidence rate ratio **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) with a lower income at a higher risk of earlier death [21]. Discussion One possible explanation to why high income is associated to improved survival may be a better performance status to This is, to our knowledge, the largest population-based study begin with, which the observed requirement of fewer pre- addressing the impact of various clinical and sociodemo- diagnostic visits could suggest. The overall mortality and graphic factors on survival and resource use in patients with morbidity in general is well known to be higher in popula- HGG. It is obvious, that sociodemographic status in addi- tions with low versus high socioeconomic status [22]. tion to patient and disease specific factors did affect survival The lack of any correlation between country of birth and and health care resource use in 845 patients diagnosed with survival are in agreement with other studies, which found no HGG during 2001–2013. A high disposable income was difference in survival for patients with astrocytoma accord- associated with a better survival, while older age, male sex ing to racial disparities [7, 23]. and high tumour grade were associated with worse survival. Field et al. published a multivariable analysis from a com- Country of birth did not show any significant correlation prehensive dataset including 542 patients with glioblastoma. with survival. They report age, poor performance status, operation type and Why females and patients with a higher disposable enrolment in clinical trial to be independent predictors for income have a better chance of survival after HGG diagnosis overall survival in multivariable analysis. In contrast to our can only be speculated on. However, our data is in line with data, they did not find that socioeconomic status, including those published by Sherwood et all, who also found patients 1 3 606 Journal of Neuro-Oncology (2018) 139:599–608 Table 4 The effect of clinical and sociodemographic factors on number of inpatient care days (univariable and multivariable regression analysis) for the Stockholm population (n = 845) Negative binomial regression Preoperative days (uni- Preoperative days (multi- Postoperative days (uni- Postoperative days (multi- variable) variable) variable) variable) IRR p value IRR p value IRR p value IRR p value Age category (ref: 18–39)  40–59 1.36 (0.90–2.05) 0.142 1.62 (1.13–2.31) 0.009 1.96 (1.49–2.59) < 0.001 2.04 (1.52–2.74) < 0.001  60–69 1.75 (1.16–2.66) 0.008 1.91 (1.31–2.79) 0.001 2.53 (1.89–3.40) < 0.001 2.40 (1.72–3.35) < 0.001  70– 2.57 (1.61–4.11) < 0.001 4.79 (3.07–7.49) < 0.001 1.65 (1.09–2.48) 0.017 1.64 (1.04–2.59) 0.033 Year of diagnosis (ref: 01/04)  05/07 0.74 (0.55–1.01) 0.06 0.69 (0.51–0.92) 0.011 1.27 (0.95–1.69) 0.101 1.33 (0.99–1.80) 0.06  08/10 0.25 (0.19–0.34) < 0.001 0.18 (0.13–0.24) < 0.001 1.38 (1.05–1.81) 0.022 1.27 (0.96–1.68) 0.089  11/13 0.25 (0.19–0.34) < 0.001 0.14 (0.10–0.19) < 0.001 1.04 (0.79–1.36) 0.795 1.06 (0.80–1.42) 0.674 Sex  Male 0.93 (0.74–1.17) 0.528 0.91 (0.74–1.13) 0.399 0.95 (0.78–1.16) 0.61 0.93 (0.76–1.14) 0.487 Comorbidity 1.72 (1.37–2.16) < 0.001 1.67 (1.34–2.08) < 0.001 1.28 (1.05–1.56) 0.015 1.07 (0.86–1.33) 0.528 Educational level (ref: elementary)  High school diploma 0.71 (0.52–0.98) 0.039 0.78 (0.58–1.05) 0.101 1.02 (0.76–1.36) 0.919 1.04 (0.78–1.40) 0.785  University degree 0.55 (0.40–0.77) < 0.001 0.84 (0.62–1.15) 0.277 1.02 (0.76–1.36) 0.901 1.09 (0.80–1.47) 0.594 Income level (ref: low)  Medium 0.84 (0.65–1.09) 0.192 0.91 (0.72–1.16) 0.451 1.16 (0.93–1.46) 0.181 1.10 (0.87–1.37) 0.433  High 0.38 (0.28–0.51) < 0.001 0.74 (0.54–1.02) 0.068 1.00 (0.77–1.31) 0.993 0.91 (0.68–1.22) 0.535 Country of birth (ref: Sweden)  Born in Europe (not Sweden) 0.97 (0.68–1.38) 0.868 1.13 (0.82–1.54) 0.463 1.00 (0.73–1.38) 0.994 1.11 (0.81–1.53) 0.505  Born outside Europe 0.90 (0.55–1.46) 0.673 1.10 (0.68–1.79) 0.690 0.84 (0.56–1.27) 0.406 0.82 (0.53–1.28) 0.376 Grade IV (ref: grade III)  Grade IV 0.94 (0.68–1.30) 0.705 1.06 (0.78–1.44) 0.709 1.41 (1.09–1.83) 0.009 1.21 (0.92–1.59) 0.172 Bold values indicate the p value ≤ 0.05 The analysis of preoperative days in hospital (1 year before diagnosis) include all 845 in the subpopulation but the postoperative days in hospital include only the 454 patients alive after 1 year *IRR (CI) = incidence rate ratio with 95% confidence intervals **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) income, had impact on survival [10]. However, the study analysis of resource use confirmed this in univariable analy - populations and included variables are not exactly the same. sis. However, in multivariable analysis comorbidities was not Our data is in line with other studies that older patients an independent prognostic factor, which may be explained have worse survival [24]. However, this study does not by its covariation with some of the sociodemographic factors include analysis of specific treatments given and it is pos- included, such as income and education. sible that one explanation to age being highly significant In addition to comorbidity, other clinical and sociodemo- is simply because older patients are more often excluded graphic factors showed significant differences in relation to from more aggressive and possibly more effective treatments resource use, before and after diagnosis. The year before for HGG [25]. Older patients and those with comorbidities HGG diagnosis, patients with a higher income level as well had twice as many prediagnostic visits before they came to as patients with a higher educational level had both fewer diagnosis with HGG. hospital admissions as well as outpatient visits. This was not The importance of comorbidities is well known in the significant at the multivariable level but may be interesting clinic situation, still little is known to what extent it affects to further investigate in the future. Of course, the situation outcome and resource use for HGG patients. Our previously is complex and multifactorial but we wanted to investigate published data, showed comorbidities to be associated with different sociodemographic factors’ possible associations decreased survival and increased resource use in patients with resource use and delayed diagnosis, which may have with primary brain tumours (not only HGG) [8]. The present impact on survival. 1 3 Journal of Neuro-Oncology (2018) 139:599–608 607 Table 5 The effect of clinical and sociodemographic factors on lead times in care process for the 845 patients in the Stockholm region diagnosed with high grade glioma 2001–2013 Negative binomial regression From diagnosis to surgery From diagnosis to surgery From surgery to start of From surgery to start of (univariate) (multivariate) non-surgical cancer treat- non-surgical cancer treat- ment (univariate) ment (multivariate) IRR (CI) p value IRR (CI) p value IRR (CI) p value IRR (CI) p value Age category (ref: 18–39)  40–59 0.59 (0.44–0.81) 0.001 1.13 (0.82–1.55) 0.455 0.42 (0.33–0.52) 0.000 0.67 (0.53–0.86) 0.002  60–69 0.55 (0.40–0.75) 0.000 1.30 (0.94–1.82) 0.117 0.40 (0.32–0.50) 0.000 0.68 (0.52–0.87) 0.003  70– 0.82 (0.57–1.20) 0.306 1.64 (1.10–2.45) 0.015 0.73 (0.52–1.02) 0.063 1.21 (0.84–1.75) 0.306 Year of diagnosis (ref: 01/04)  05/07 1.57 (1.03–2.38) 0.035 0.91 (0.61–1.35) 0.635 0.43 (0.29–0.62) 0.000 0.64 (0.42–0.98) 0.038  08/10 1.40 (0.93–2.10) 0.105 1.36 (0.92–2.01) 0.127 0.26 (0.18–0.37) 0.000 0.42 (0.27–0.63) 0.000  11/13 0.91 (0.61–1.37) 0.651 0.97 (0.65–1.43) 0.870 0.29 (0.20–0.42) 0.000 0.40 (0.26–0.61) 0.000 Sex  Male 0.83 (0.69-1.00) 0.049 0.91 (0.76–1.08) 0.286 1.25 (1.08–1.46) 0.003 1.01 (0.87–1.18) 0.863 Comorbidity* 1.57 (1.31–1.89) 0.000 1.53 (1.28–1.83) 0.000 0.81 (0.70–0.95) 0.010 0.94 (0.81–1.10) 0.460 Educational level (ref: elementary)  High school diploma 1.00 (0.78–1.29) 0.997 1.11 (0.87–1.41) 0.409 1.19 (0.96–1.48) 0.114 1.06 (0.86–1.31) 0.588  University degree 0.74 (0.57–0.95) 0.018 1.02 (0.79–1.31) 0.879 1.32 (1.07–1.64) 0.011 0.98 (0.78–1.22) 0.843 Income level (ref: low)  Medium 1.30 (1.06–1.60) 0.010 1.04 (0.86–1.27) 0.669 0.68 (0.57–0.82) 0.000 0.87 (0.73–1.04) 0.122  High 0.83 (0.66–1.05) 0.116 0.88 (0.69–1.11) 0.269 0.93 (0.77–1.13) 0.467 1.10 (0.90–1.34) 0.355 Country of birth (ref: Sweden)  Born in Europe (not Sweden) 1.08 (0.82–1.41) 0.000 0.70 (0.55–0.91) 0.006 0.80 (0.63–1.02) 0.077 0.94 (0.76–1.16) 0.565  Born outside Europe 3.43 (2.41–4.89) 0.000 2.79 (1.91–4.09) 0.000 0.78 (0.56–1.09) 0.148 0.70 (0.50–0.97) 0.030 Grade IV (ref: grade III)  Grade IV 0.37 (0.29–0.49) 0.000 0.48 (0.36–0.63) 0.000 0.50 (0.40–0.63) 0.077 0.70 (0.55–0.89) 0.003 Bold values indicate the p value ≤ 0.05 *IRR (CI) = incidence rate ratio with 95% confidence intervals **Any comorbidity (any comorbidity or no comorbidity according to the Elixhauser definition) Patients born in Europe, but not in Sweden, had signifi- Another limitation of our paper is that treatment regi- cantly more postoperative outpatient visits. One can specu- men is not included in the analyses, as this stratification late if this may be explained by difficulties with language would render too small subgroups for robust statistical and communication as information, written information is analyses. But we do know that treatment affect survival. usually in Swedish, or if it is due to differences in culture, We also lack information about other factors known to but this needs to be further studied. affect survival like performance status, extent of resection Our analysis of lead times in the care process was and molecular subtypes. On the other hand, one of the employed in order to investigate potential differences strengths is that the analysis includes all patients diag- between patients with different clinical and or/sociode- nosed with HGG between January 1st 2001 and December mographic status. Country of birth did affect time from 31st 2013, in one large region. The unique administrative diagnosis to surgery as well as time from surgery to start database, which covers hospital admissions as well as out- of non-surgical cancer treatment. However, the number of patient visits to specialists and visits in primary care, was patients born in other countries than Sweden, especially out- linked to several registries by the personal identification side Europe, is relatively limited which make it difficult to number. make firm conclusions. And of course, differences in a few Over time we found less need for outpatient visits the days may not be of clinical importance and further studies year before HGG diagnosis but more outpatient visits the are needed to see if the number and extent of differences year after, which can be explained by improved pre diagnos- stays the same. tic care and more interventions after diagnosis. In addition, time from surgery to start of non-surgical cancer treatment 1 3 608 Journal of Neuro-Oncology (2018) 139:599–608 outcomes among elderly individuals with primary malignant decreased over time and was significantly shorter during the astrocytoma. J Neurosurg 108(4):642–648 last half of the study period. 8. Bergqvist J, Iderberg H, Mesterton J, Bengtsson N, Wettermark B, The effect of sociodemographic factors on survival could Henriksson R (2017) Healthcare resource use, comorbidity, treat- potentially be reduced with increased awareness of these ment and clinical outcomes for patients with primary intracranial tumors: a Swedish population-based register study. Acta Oncol inequalities and relocation of resources in order to balance 56(3):405–414 inequities when necessary. For example, we need to increase 9. Danzig MR, Weinberg AC, Ghandour RA, Kotamarti S, McKi- knowledge of when to address the health care system among ernan JM, Badani KK (2014) The association between socioeco- all people regardless of sociodemographic status, like sug- nomic status, renal cancer presentation, and survival in the United States: a survival, epidemiology, and end results analysis. Urology gested by Whitaker et al. [4]. 84(3):583–589 In conclusion, sociodemographic factors have impact on 10. Field KM, Drummond KJ, Yilmaz M, Tacey M, Compston D, survival and resource use for patients with HGG and with Gibbs P et al (2013) Clinical trial participation and outcome for increased awareness and further studies we can reduce the patients with glioblastoma: multivariate analysis from a compre- hensive dataset. J Clin Neurosci 20(6):783–789 inequalities and further reduce clinically relevant differences. 11. Hastert TA, Beresford SA, Sheppard L, White E (2015) Dispari- ties in cancer incidence and mortality by area-level socioeconomic Acknowledgements The authors would like to thank Nils Bengtsson status: a multilevel analysis. J Epidemiol Community Health at Ivbar Institute for work with the database. 69(2):168–176 12. Aizer AA, Ancukiewicz M, Nguyen PL, Shih HA, Loeffler JS, Compliance with ethical standards Oh KS (2014) Underutilization of radiation therapy in patients with glioblastoma: predictive factors and outcomes. Cancer 120(2):238–243 Conflict of interest None to declare for Jenny Bergqvist and Roger 13. Fritz AG (2013) International classification of diseases for oncol- Henriksson. Hanna Iderberg and Johan Mesterton are employed at ogy: ICD-O, 3rd edn, 1st rev. edn. World Health Organization, Ivbar Institute, a research company specialized in health care govern- Geneva, p. viii, 242 ance and analysis of health care data. Johan Mesterton is a shareholder 14. Barlow L, Westergren K, Holmberg L, Talback M (2009) The at Ivbar Institute. completeness of the Swedish Cancer Register: a sample survey for year 1998. Acta Oncol 48(1):27–33 Open Access This article is distributed under the terms of the Crea- 15. National Board of Health and Welfare (2013) The Swedish Cancer tive Commons Attribution 4.0 International License (http://creat iveco Registry (database) mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 16. Statistics Sweden: longitudinal integration database for health tion, and reproduction in any medium, provided you give appropriate insurance and labour market studies (LISA by Swedish acronym) credit to the original author(s) and the source, provide a link to the (2016) Creative Commons license, and indicate if changes were made. 17. The National Board of Health and Welfare (2016) Causes of Death (database) 18. Carlsson AC, Wandell P, Osby U, Zarrinkoub R, Wettermark B, Ljunggren G (2013) High prevalence of diagnosis of diabetes, References depression, anxiety, hypertension, asthma and COPD in the total population of Stockholm, Sweden—a challenge for public health. 1. Asklund T, Malmstrom A, Bergqvist M, Bjor O, Henriksson R BMC Public Health 13:670 (2015) Brain tumors in Sweden: data from a population-based 19. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, registry 1999–2012. Acta Oncol 54(3):377–384 Jouvet A et al (2007) The 2007 WHO classification of tumours of 2. Bauchet L, Mathieu-Daude H, Fabbro-Peray P, Rigau V, Fabbro the central nervous system. Acta Neuropathol 114(2):97–109 M, Chinot O et al (2010) Oncological patterns of care and out- 20. Elixhauser A, Steiner C, Harris DR, Coffey RM (1998) Comor - come for 952 patients with newly diagnosed glioblastoma in 2004. bidity measures for use with administrative data. Med Care Neuro Oncol 12(7):725–735 36(1):8–27 3. Lacroix M, Abi-Said D, Fourney DR, Gokaslan ZL, Shi W, 21. Sherwood PR, Dahman BA, Donovan HS, Mintz A, Given CW, DeMonte F et al (2001) A multivariate analysis of 416 patients Bradley CJ (2011) Treatment disparities following the diagnosis with glioblastoma multiforme: prognosis, extent of resection, and of an astrocytoma. J Neurooncol 101(1):67–74 survival. J Neurosurg 95(2):190–198 22. Chetty R, Stepner M, Abraham S, Lin S, Scuderi B, Turner N et al 4. Whitaker KL, Scott SE, Wardle J (2015) Applying symptom (2016) The Association between income and life expectancy in appraisal models to understand sociodemographic differences in the United States, 2001–2014. JAMA 315(16):1750–1766 responses to possible cancer symptoms: a research agenda. Br J 23. Robertson JT, Gunter BC, Somes GW (2002) Racial differences Cancer 112(Suppl 1):S27-34 in the incidence of gliomas: a retrospective study from Memphis, 5. Chakrabarti I, Cockburn M, Cozen W, Wang YP, Preston- Tennessee. Br J Neurosurg 16(6):562–566 Martin S (2005) A population-based description of glioblas- 24. Thumma SR, Fairbanks RK, Lamoreaux WT, Mackay AR, Dema- toma multiforme in Los Angeles County, 1974–1999. Cancer kas JJ, Cooke BS et al (2012) Effect of pretreatment clinical fac- 104(12):2798–2806 tors on overall survival in glioblastoma multiforme: a surveillance 6. Khanolkar AR, Ljung R, Talbäck M, Brooke HL, Carlsson S, epidemiology and end results (SEER) population analysis. World Mathiesen T, Feychting M (2016) Socioeconomic position and the J Surg Oncol 10:75 risk of brain tumour: a Swedish national population-based cohort 25. Tanaka S, Meyer FB, Buckner JC, Uhm JH, Yan ES, Par- study. J Epidemiol Community Health ney IF (2013) Presentation, management, and outcome of 7. Barnholtz-Sloan JS, Williams VL, Maldonado JL, Shahani D, newly diagnosed glioblastoma in elderly patients. J Neurosurg Stockwell HG, Chamberlain M et al (2008) Patterns of care and 118(4):786–798 1 3

Journal

Journal of Neuro-OncologySpringer Journals

Published: May 30, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off