Reappraisal of the Inverse Care Law – Analysis of Results 1

Part 1 – Quantitative Analysis of Socio-economic Status and Age Group, with Experience of Health Services

This section includes the main results in the form of contingency tables, statistics and bar charts. Some of the results in the contingency tables are highlighted where there is a dramatic difference in expected count and actual count. For a list of all variables and original values see Appendix 4.  An unusual distribution was found for ‘age left education’ (see Appendix 5), therefore the distribution of age left education and age group was presented in bar charts and analysed (See Appendix 5) which revealed a large proportion (59.8%) missing for ‘age left education’, explaining the overrepresentation of those that left school age 16 or less; it is commented that those missing are likely to be aged 16-35 (See Appendix 5). Age was then analysed in relation to age left education. As respondents still in full-time education were so highly concentrated in the age group 16-35 (See Appendix 6) and because this category was not useful for revealing socioeconomic status, those still in full-time education were filtered out and age and age left education re-examined.

Analysis of Results Part 1 – Socioeconomic Status and Health Care Experience

Legend for Contingency Tables:

  • + = higher than expected count
  • – = lower than expected count
  • / = little difference between actual and expected count

Table I. Contingency Table Age Group and Age Left Full-Time Education, Filtering Out Respondents Still in Full-Time Education

Age group, taken from response or from sample information if response missing
Age left education 16-35 years of age 36-50 years of age 51-65 years of age Over 65 years of age Row Total
16 or less count 503- 1205- 1882+ 1880 + 5470
16 or less expected count 933.9 1426.1 1706.5 1403.5 5470.0
17 or 18 count 461+ 621+ 521- 312- 1915
17 or 18 expected count 327.0 499.2 597.4 491.4 1915.0
19 or over count 651 + 640+ 548- 235 – 2074
19 or over expected count 354.1 540.7 647.0 532.1 2074.0
Column Total 1615 2466 2951 2427 9459

The biggest differences in expected and actual count were between the youngest and oldest age groups, at the highest and lowest levels of education and fit the expected pattern: those that left school at 16 or less were overrepresented by the age groups 50 and over.

Table II. Chi Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 1021.056a 6 .000
Likelihood Ratio 1043.165 6 .000
Linear-by-Linear Association 950.882 1 .000
N of Valid Cases 9459
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 326.96.

It is significant and valid. As the obtained value of 1021.056 is much higher than the critical value of 12.59 it suggests there is a strong correlation between age group and age left education.

Given that age has been shown to affect school-leaving age, answers to questions were analysed in relation to both age group and age left education, to consider how this relationship interfered with the association between socio-economic status and health care service received.

Indicator 1 – Time Waited for an Appointment

Table III. Contingency Table Age Left Education and Time Waited for Appointment

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
How old were you when you left full-time education? * The last time you saw a doctor from your GP practice/ health centre, how long did you wait for an appointment? 6761 71.5% 2698 28.5% 9459 100.0%

How long did you wait for an appointment?

Age left education I was seen without an appointment I was seen on the same working day I waited 1 or 2 working days I waited more than 2 working days Row total
16 or less count 180+ 1607/ 1271+ 893- 3951
16 or less expected count 167.7 1605.9 1237.1 940.3 3951.0
17 or 18 count 39- 554/ 409- 367+ 1369
17 or 18 expected count 58.1 556.4 428.7 325.8 1369.0
19 or over count 68/ 587/ 437- 349/ 1441
19 or over expected count 61.2 585.7 451.2 342.9 1441.0
Column total 287 2748 2117 1609 6761

Those that left school at 16 or less had a higher than expected count for being seen without an appointment and waiting one or two days, and lower than expected count for waiting more than two working days and the opposite is the case for those that left school at 17 or 18. Whilst this pattern is contrary to expected there is little actual difference in the expected counts and actual counts.

Table IV. Chi-Square Tests
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 17.933a 6 .006
Likelihood Ratio 18.560 6 .005
Linear-by-Linear Association 1.226 1 .268
N of Valid Cases 6761
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 58.11.

Results are valid and significant. The obtained value is slightly higher than the critical value 12.59, suggesting there is not a particularly strong correlation between age left education and how long patients waited for an appointment.

Table V. Contingency Table for Age Group and Time Waited for Appointment

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
Age group, taken from response or from sample information if response missing * The last time you saw a doctor from your GP practice/ health centre, how long did you wait for an appointment? 6761 71.5% 2698 28.5% 9459 100.0%

How long did you wait for an appointment?

Age Group I was seen without an appointment I was seen on the same working day I waited 1 or 2 working days I waited more than 2 working days Row total
16-35 years count 45/ 506+ 375/ 270- 1196
16-35 years expected count 50.8 486.1 374.5 284.6 1196.0
36-50 years count 64- 741+ 515- 463+ 1783
36-50 years expected count 75.7 724.7 558.3 424.3 1783.0
51-65 years count 78/ 802- 664+ 489/ 2033
51-65 years expected count 86.3 826.3 636.6 483.8 2033.0
Over 65 years  count 100+ 699- 563+ 387- 1749
Over 65 years expected count 74.2 710.9 547.6 416.2 1749.0
Column Total 287 2748 2117 1609 6761

There is no clear pattern.

Table VI. Chi Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 25.643a 9 .002
Likelihood Ratio 24.937 9 .003
Linear-by-Linear Association .844 1 .358
N of Valid Cases 6761
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 50.77.

It is significant and valid. The obtained value of 25.643 is greater than the critical value of 16.92, however this does not suggest a strong correlation and it is not clear in which direction.

Indicator 2 – Whether Given Enough Time to Discuss Medical Problem

Table VII. Contingency Table for Education Leaving Age and Time Given to Discuss Medical Problem

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
How old were you when you left full-time education? * Were you given enough time to discuss your health or medical problem with the doctor? 7986 84.4% 1473 15.6% 9459 100.0%

Were you given enough time to discuss your problem?

Age left education Yes, definitely Yes, to some extent No Row Total
16 or less count 3579+ 951- 130- 4660
16 or less expected count 3514.0 986.7 159.3 4660.0
17 or 18 count 1177- 353+ 68+ 1598
17 or 18 expected count 1205.0 338.4 54.6 1598.0
19 or over count 1266- 387+ 75+ 1728
19 or over expected count 1303.0 365.9 59.1 1728.0
Column total 6022 1691 273 7986

The pattern is the opposite than expected; those that left school the earliest were more likely to have been given enough time to discuss their medical problem.

Table VIII. Chi Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 19.009a 4 .001
Likelihood Ratio 18.786 4 .001
Linear-by-Linear Association 15.126 1 .000
N of Valid Cases 7986
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 54.63.

It is both significant and valid. The obtained value of 19.009 is higher than the critical value of 9.49; there appears to be a slight relationship between age left education and whether patients are given enough time to discuss their problem.

Table IX. Contingency Table for Age Group and Time Given to Discuss Medical Problem

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
Age group, taken from response or from sample information if response missing * Were you given enough time to discuss your health or medical problem with the doctor? 7986 84.4% 1473 15.6% 9459 100.0%

Were you given enough time to discuss your problem?

Age Group Yes, definitely Yes, to some extent No Row Total
16-35 years count 917- 361+ 74+ 1352
16-35 years expected count 1019.5 286.3 46.2 1352.0
36-50 years count 1496- 489+ 82+ 2067
36-50 years expected count 1558.7 437.7 70.7 2067.0
51-65 years count 1864+ 494- 79/ 2437
51-65 years expected count 1837.7 516.0 83.3 2437.0
Over 65 years  count 1745+ 347- 38- 2130
Over 65 years expected count 1606.2 451.0 72.8 2130.0
Column Total 6022 1691 273 7986

There is a clear pattern the opposite to expected: the younger age group are less likely to be given enough time to discuss the problem and more likely to be given enough time ‘to some extent’ or not given enough time at all, whilst it is the opposite for older groups of respondents.

Table X. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 111.040a 6 .000
Likelihood Ratio 112.354 6 .000
Linear-by-Linear Association 109.535 1 .000
N of Valid Cases 7986
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 46.22.

It is significant and valid. The obtained value of 111.040 is much higher than the critical value of 12.59, which suggests there is a strong association between age group and whether patients are given enough time to discuss their medical problem.

Indicator 3 – Whether Treated with Respect and Dignity

Table XI. Contingency Table of Age Left Education and Whether Treated with Respect and Dignity

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
How old were you when you left full-time education? * Did the doctor treat you with respect and dignity? 8016 84.7% 1443 15.3% 9459 100.0%

 

Did the doctor treat you with respect and dignity?

 
Age left education Yes, all of the time Yes, some of the time No Row Total
16 or less count 4379+ 265- 42- 4686
16 or less expected count 4330.6 295.2 60.2 4686.0
17 or 18 count 1453- 120+ 29/ 1602
17 or 18 expected count 1480.5 100.9 20.6 1602.0
19 or over count 1576- 120+ 32+ 1728
19 or over expected count 1596.9 108.9 22.2 1728.0
Column total 7408 505 103 8016

There is a slight pattern the opposite to expected; those that left education at 16 or less had a higher than expected count for being treated with respect and dignity and those that left school aged 17 or over had a lower than expected count.

Table XII. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 22.435a 4 .000
Likelihood Ratio 22.046 4 .000
Linear-by-Linear Association 16.718 1 .000
N of Valid Cases 8016
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 20.58.

It is valid and significant. The obtained value of 22.435 is higher than 9.49; there is a slight correlation between age left education and whether patients were treated with respect and dignity.

Table XIII. Contingency Table of Age Group and Whether Treated with Respect and Dignity

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
Age group, taken from response or from sample information if response missing * Did the doctor treat you with respect and dignity? 8016 84.7% 1443 15.3% 9459 100.0%

Did the doctor treat you with respect and dignity?

Age Group Yes, all of the time Yes, some of the time No Row Total
16-35 years count 1178- 143+ 37+ 1358
16-35 years expected count 1255.0 85.6 17.4 1358.0
36-50 years count 1857- 175+ 41+ 2073
36-50 years expected count 1915.8 130.6 26.6 2073.0
51-65 years count 2289+ 133- 19- 2441
51-65 years expected count 2255.9 153.8 31.4 2441.0
Over 65 years  count 2084+ 54- 6- 2144
Over 65 years expected count 1981.4 135.1 27.5 2144.0
Column Total 7408 505 103 8016

There is a pattern the opposite to expected; the older age groups had a higher than expected count for being treated by the doctor with respect and dignity and the opposite is the case for younger age groups ’16-35’ and ’36-50’.

Table XIV. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 168.848a 6 .000
Likelihood Ratio 177.726 6 .000
Linear-by-Linear Association 162.996 1 .000
N of Valid Cases 8016
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 17.45.

It is significant and valid. The obtained value of 168.848 is much higher than the critical value of 12.59, therefore there is a strong association between age and whether patients were treated with respect and dignity.

Indicator 4 – Whether Been Referred to a Specialist In the Last Twelve Months

Table XV. Contingency Table for Age Left Education and Whether Been Referred to Specialist

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
How old were you when you left full-time education? * In the last 12 months, has anyone at your GP practice/ health centre referred you to a specialist (e.g. a hospital consultant)? 9298 98.3% 161 1.7% 9459 100.0%

In the last 12 months has anyone at your GP referred you to a specialist?

Age left education Yes

No

Row Total
16 or less count 2002+ 3357- 5359
16 or less expected count 1925.0 3434.0 5359.0
17 or 18 count 655- 1232+ 1887
17 or 18 expected count 677.8 1209.2 1887.0
19 or over count 683- 1369+ 2052
19 or over expected count 737.1 1314.9 2052.0
Column total 3340 5958 9298

The pattern is the opposite to expected: those that left education age 16 or less had a higher than expected count for being referred to a specialist in the last 12 months; those that left education aged 17 or over had a lower than expected count.

Table XVI. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 12.202a 2 .002
Likelihood Ratio 12.252 2 .002
Linear-by-Linear Association 11.969 1 .001
N of Valid Cases 9298
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 677.84.

It is significant and valid. The obtained value of 12.202 is slightly higher than the critical value of 5.99, it suggests there is a slight association between age left education and referral to a specialist.

Table XVII. Contingency Table for Age Group and Whether Been Referred to a Specialist

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
Age group, taken from response or from sample information if response missing * In the last 12 months, has anyone at your GP practice/ health centre referred you to a specialist (e.g. a hospital consultant)? 9817 40.5% 14430 59.5% 24247 100.0%

It is worth noting that the percentage missing for this particular question in relation to age group was very high (59.5%). This may have affected results. As suggested previously the greatest proportion missing is likely to be from the 16-35 age group.

 

In the last 12 months has anyone at your GP referred you to a specialist?

Age Group Yes

No

Row Total
16-35 years count 566- 1310+ 1876
16-35 years expected count 670.6 1205.4 1876.0
36-50 years count 889/ 1590/ 2479
36-50 years expected count 886.1 1592.9 2479.0
51-65 years count 1009- 1969+ 2978
51-65 years expected count 1064.5 1913.5 2978.0
Over 65 years  count 1045+ 1439- 2484
Over 65 years expected count 887.9 1596.1 2484.0
Column Total 3509 6308 9817

There is a slight pattern at opposite ends of the age spectrum: those aged 16-35 had a lower than expected count for being referred to a specialist and those aged 65 or over had a higher than expected count.

Table XVIII. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 73.154a 3 .000
Likelihood Ratio 72.884 3 .000
Linear-by-Linear Association 52.072 1 .000
N of Valid Cases 9817
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 670.56.

It is significant and valid. The obtained value of 73.145 is much greater than the critical value of 7.82, therefore there is an association between age group and referral to a specialist.

Indicator 5 – Was the Main Reason Dealt With to Your Satisfaction?

Table XIX. Contingency Table for Age Left Education and Whether Problem Dealt With to Satisfaction

Case Processing Summary
Cases
Valid Missing Total
N Percent N Percent N Percent
How old were you when you left full-time education? * Was the main reason you went you your GP practice/ health centre dealt with to your satisfaction? 9179 97.0% 280 3.0% 9459 100.0%
Was the main reason you went to your GP dealt with to your satisfaction?
Age left education Yes, completely Yes, to some extent No Row Total
16 or less count 4071+ 1096- 152- 5319
16 or less expected count 3979.3 1167.1 172.7 5319.0
17 or 18 count 1361- 432+ 63/ 1856
17 or 18 expected count 1388.5 407.2 60.3 1856.0
19 or over count 1435- 486+ 83+ 2004
19 or over expected count 1499.2 439.7 65.1 2004.0
Column total 6867 2014 298 9179

Results are the opposite to expected: those that left school earlier (16 or less) had a higher than expected count for having their problem dealt with to their satisfaction; those that left aged 19 or over had a lower than expected count.

Table XX. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 23.669a 4 .000
Likelihood Ratio 23.298 4 .000
Linear-by-Linear Association 23.029 1 .000
N of Valid Cases 9179
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 60.26.

It is significant and valid. The obtained value 23.669 is greater than the critical value of 9.49; there appears to be a slight correlation between age left education and having the problem deal with satisfactorily.

Table XXI. Contingency Table for Age Group and Whether Problem Dealt With to Satisfaction

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
Age group, taken from response or from sample information if response missing * Was the main reason you went you your GP practice/ health centre dealt with to your satisfaction? 9179 97.0% 280 3.0% 9459 100.0%

Was the main reason you went to your GP dealt with to your satisfaction?

Age Group Yes, completely Yes, to some extent No Row Total
16-35 years count 1065- 437+ 78+ 1580
16-35 years expected count 1182.0 346.7 51.3 1580.0
36-50 years count 1653- 631+ 103+ 2387
36-50 years expected count 1785.8 523.7 77.5 2387.0
51-65 years count 2183+ 606- 75- 2864
51-65 years expected count 2142.6 628.4 93.0 2864.0
Over 65 years  count 1966+ 340- 42- 2348
Over 65 years expected count 1756.6 515.2 76.2 2348.0
Column Total 6867 2014 298 9179

There is a clear pattern and a large difference between expected and actual counts though it is the opposite than expected. The younger age groups of 16-35 and 36-50 had a higher than expected count for only having their problem dealt with ‘to some extent’ or not at all; older age groups 51-65 and 65+ had a higher than expected count for having their problem dealt with ‘completely’ to their satisfaction.

Table XXII. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 194.196a 6 .000
Likelihood Ratio 199.492 6 .000
Linear-by-Linear Association 178.098 1 .000
N of Valid Cases 9179
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 51.30.

It is significant and valid. The obtained value of 194.196 is much greater than the critical value 12.59; therefore there is a strong association between age group and whether or not the main problem was dealt with satisfactorily.

Indicator 6 – Whether or Not Been Put Off Going to GP Due to Inconvenient Opening Times

Table XXIII. Contingency Table for  Age Left Education and Whether Been Put Off Appointments Due to Inconvenient Opening Times

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
How old were you when you left full-time education? * In the last 12 months, have you ever been put off going to your GP practice/ health centre because the opening times are inconvenient for you? 9286 98.2% 173 1.8% 9459 100.0%
 

Have you been put off going to your GP because of inconvenient opening times?

 
Age left education Yes, often Yes, sometimes No Row Total
16 or less count 282- 770- 4314+ 5366
16 or less expected count 371.6 924.6 4069.9 5366.0
17 or 18 count 175+ 388+ 1322- 1885
17 or 18 expected count 130.5 324.8 1429.7 1885.0
19 or over count 186+ 442+ 1407- 2035
19 or over expected count 140.9 350.6 1543.5 2035.0
Column total 643 1600 7043 9286

There is a strong pattern: Those that left education aged 16 or less had a lower than expected count for being put off going to their GP because of inconvenient opening times and those that left school at 17 or 18 or 19 had a higher than expected count.

Table XXIV. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 147.941a 4 .000
Likelihood Ratio 146.423 4 .000
Linear-by-Linear Association 121.274 1 .000
N of Valid Cases 9286
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 130.52.

It is valid and significant. The obtained value of 147.941 is much greater than the critical value of 9.49, suggesting there is a strong relationship between age left education and whether or not they are likely to be put off going to their GP due to inconvenient opening times.

Table XXV. Contingency Table for Age Group and Whether Been Put Off Making an Appointment Due to Inconvenient Opening Times

Case Processing Summary

Cases

Valid Missing Total
N Percent N Percent N Percent
Age group, taken from response or from sample information if response missing * In the last 12 months, have you ever been put off going to your GP practice/ health centre because the opening times are inconvenient for you? 9286 98.2% 173 1.8% 9459 100.0%

Have you been put off going to your GP because of inconvenient opening times?

 
Age Group Yes, often Yes, sometimes No Row Total
16-35 years count 196+ 388 + 1006 – 1590
16-35 years expected count 110.1 274.0 1205.9 1590.0
36-50 years count 238+ 538+ 1651 – 2427
36-50 years expected count 168.1 418.2 1840.8 2427.0
51-65 years count 176- 509+ 2215+ 2900
51-65 years expected count 200.8 499.7 2199.5 2900.0
Over 65 years  count 33 – 165 – 2171 + 2369
Over 65 years expected count 164.0 408.2 1796.8 2369.0
Column Total 643 1600 7043 9286

The results show a pattern opposite to expectations:  the younger age groups (16-35 and 36-50) had a higher count than expected for being put off from making an appointment either ‘often’ or ‘sometimes’ whereas those aged 65 or over had a much lower than expected count for being put off due to inconvenient opening times.

Table XXVI. Chi-Square Statistic
Value df Asymp. Sig. (2-sided)
Pearson Chi-Square 561.500a 6 .000
Likelihood Ratio 629.109 6 .000
Linear-by-Linear Association 505.170 1 .000
N of Valid Cases 9286
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 110.10.

It is valid and significant. The obtained value of 561.500 is much greater than the critical value of 12.59, therefore there is likely to be a very strong association between age group and whether put off due to inconvenient opening times.