Table 1. Characteristics of included studies potentially suitable for meta-analysis (universal health care systems).

Paper	Country of Study	Data Source (s)	Population Included	Years of Diagnosis	Measure of SEP	No. of SEP groups	Treatment given within	Age Range	Confounders Controlled For:					Quality Score
									Age	Sex	Stage	Histology	Other
Berglund et al, 2010 [19]	Sweden	Regional Lung Cancer Register (RLCR) - Sweden, Cause of Death Register and LISA (insurance and demographics)	Uppsala/Orebro region in central Sweden	1996–2004	Education levela	3	NR	30+	Yes	Yes	Yes	Yes	Performance status, year of diagnosis, smoking status	6
Berglund et al, 2012 [22]	England	Thames Cancer Registry, HES, LUCADA	South-east England	2006–2008	IMD 2007 income domain	5	NR	0–80+	Yes	Yes	Yes	Yes	Co-morbidity	6
Campbell et al, 2002 [35]	Scotland	Scottish Cancer Registry and hospital case notes	Random sample from North/NE Scotland (with hospital record)	1995–1996	Carstairs Index	5	12 months	NR	Yes	Yes	Yes	Yes	Health board, distance to cancer centre, mode of admission	5
Crawford et al, 2009 [36]	England	Northern and Yorkshire Cancer Registry and Information Service (NYCRIS)	Northern and Yorkshire region	1994–2002	IMD 2004 (access to services domain removed)	4	6 months	NR	Yes	Yes	No	Yes	Travel time (but overall results not stratified by travel time used here). Histology not included in receipt of any treatment analysis.	4
Erridge et al, 2002 [37]	Scotland	Scottish Cancer Registry and medical records	Scotland (with hospital record)	1995	Carstairs Index	5	6 months	<60– 80+	Yes	Yes	Yes	Yes	Health board (not inc in receipt of radiotherapy), diagnosis by specialist, management by oncologist	6
Erridge et al, 2009 [18]	Scotland/Canada	Scottish Cancer Registry and medical records; British Columbia Cancer Registry	Scotland/British Columbia	1995	Carstairs Index/average household income	2	6 months	<60– 80+	Yes	Yes	Yes	Yes	Travel time, CT scan	4
Gregor et al, 2001 [38]	Scotland	Scottish Cancer Registry and medical records	Scotland (with hospital record)	1995	Carstairs Index	5	6 months	<60–80+	Yes	Yes	Yes	Yes	Referral to specialist within 6 months of diagnosis	6
Jack et al, 2003 [39]	England	Thames Cancer Registry	South-east England	1995–1999	Townsend (median score per health authority)	Contin-uousb	NR	<35–85+	Yes	Yes	Yes	Yes	First hospital visited is a radiotherapy centre, basis of diagnosis, incidence. Health authority/hospital used as 2nd level in multi-level model.	4
Jack et al, 2006 [40]	England	Thames Cancer Registry and medical records	South-east London (with hospital record)	1998	IMD 2000	5	6 months	<55–85+	Yes	Yes	Yes	Yes	Consultant specialty, basis of diagnosis (hospital, number of symptoms in some analyses)	6
Jones et al,2008 [41]	England	Northern and Yorkshire Cancer Registry and Information Service (NYCRIS)	Northern and Yorkshire region	1994–2002	IMD 2004 (access to services domain removed)	Contin-uousc	NR	NR	Yes	Yes	No	Yes	Travel time to hospital	4
Mahmud et al, 2003 [42]	Ireland	National Cancer Registry of Ireland (NCRI)	Republic of Ireland	1994–1998	SAHRU area-based material deprivation index	3	6 months	15–80+	Yes	Yes	No	Yes	Health board, year of diagnosis	4/2d
McMahon et al, 2011 [43]	England	Eastern Cancer Registry and Information Centre (ECRIC)	East of England	1995–2006	IMD 2004 (access to services domain removed)	5	NR	<60–80+	Yes	Yes	No	Yes	Year of diagnosis	4
Pollock &Vickers, 1998 [44]	England	HES FCEs	North/South Thames (admitted to hospital)	1992–1995	Townsend	10	NR	<100	Yes	Yes	No	No	Hospital, mode of admission	3
Raine et al, 2010 [45]	England	HES FCEs	England (admitted to hospital)	1999–2006	IMD	5	NR	50– 90+	Yes	Yes	No	No	Trust, year of admission, mode of admission	3
Riaz et al, 2012 [34]	England	NCIN/UKACR cancer registries	England	2004–2006	IMD 2004	5	NR	0– 85+	Yes	Yes	No	No	Government Office Region	4
Rich et al, 2011(1) [46]	England	LUCADA supplied by 157 NHS trusts	England	2004–2007	Townsend	5	NR	NR	Yes	Yes	Yes	Yes	Performance status. Adjusted for clustering by NHS trust	5
Rich et al, 2011(2) [21]	England	LUCADA and HES	England	2004–2008	Townsend	5	NR	30–100	Yes	Yes	Yes	Yes	Co-morbidity, ethnicity, surgery centre, radiotherapy centre, trial entry. Adjusted for clustering by NHS trust	5
Stevens et al, 2007 [23]	New Zealand	Regional hospital and oncology databases checked against NZ cancer registry	Auckland-Northland region patients managed in secondary care	2004	NZ Deprivation Index	2	NR	<60–80+	Yes	Yes	Yes	Yes	Co-morbidity, private sector care, care discussed at MDM	3
Stevens et al, 2008 [47]	New Zealand	Regional hospital and oncology databases checked against NZ cancer registry	Auckland-Northland region patients managed in secondary care	2004	NZ Deprivation Index	10	NR	<60–80+	Yes	Yes	Yes	Yes	Co-morbidity, private sector care, ethnicity	5

Quality score ranges from 1 (lowest quality) to 6 (highest quality).

^aSocioeconomic index (SEI) and household income also measured but individual education level used in analyses as it contained least missing data.

^bOdds ratio for 1 unit increase in deprivation score, range unknown.

^cOdds ratio for 1 unit increase in deprivation score, range 1–80.

^dQuality score 4 where adjusted OR used and 2 where unadjusted rates used.

HES, Hospital Episode Statistics; HES FCE, Hospital Episode Statistics Finished Consultant Episode; IMD, Index of Multiple Deprivation; LUCADA, Lung Cancer Audit; MDM, multi-disciplinary meeting; NCIN/UKACR, National Cancer Information Network/UK Association of Cancer Registries; NR, not reported; OR, odds ratio; SEP, socioeconomic position; UHCS, universal health care system.