“h1n1 by the numbers” looking at h1n1 outcomes among first nations using health records

26
1 “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records UBC Learning Circle Drs. Evan Adams & Shannon Waters April 26, 2012

Upload: liuz

Post on 14-Jan-2016

29 views

Category:

Documents


0 download

DESCRIPTION

“H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records. UBC Learning Circle Drs. Evan Adams & Shannon Waters April 26, 2012. Background. 1 st wave of H1N1 began in spring 2009 & 2 nd wave in fall 2009 BC had a unique, Tripartite response to H1N1 - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

1

“H1N1 by the numbers”Looking at H1N1 outcomes among First

Nations using health records

UBC Learning Circle

Drs. Evan Adams & Shannon Waters

April 26, 2012

Page 2: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

2

Background• 1st wave of H1N1 began in spring 2009 & 2nd

wave in fall 2009• BC had a unique, Tripartite response to H1N1

– Tripartite First Nations (FN) H1N1 Working Group with partners from FN Inuit Health Branch, Ministry of Health, Health Authorities, BC Centre for Disease Control & FN Health Council staff

• Approaches to help FN deal with H1N1– Prepositioning of antivirals (Tamiflu) in select FN

communities

Page 3: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

3

Objectives of Learning Circle

•Methods for evaluating Tripartite Response – quantitative analysis

•What can health records tell us about how FN did during H1N1?

•What do these things mean?

•What recommendations can we have from the evaluation?

Page 4: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

4

Questions for H1N1 Evaluation

• Did FN get sicker than other British Columbians?– Did influenza & pneumonia rates vary between FN & other

residents across BC? • Did having access to prepositioned Tamiflu reduce the

rates of influenza & pneumonia? • Were influenza & pneumonia rates different between

those with & without chronic conditions? • Did Tamiflu prescription rates vary between FN & other

residents across the province?

Page 5: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

5

Methods: Cohort

• Retrospective cohort analysis of BC residents between January 1, 2009 & March 31, 2010

• Study population divided into 6 groups:– Status Indians

1. Living in Local Health Areas (LHAs) where all FN communities had access to prepositioned Tamiflu (“Prepositioned LHAs”)

2. Living in LHAs where some FN communities had access to prepositioned Tamiflu (“Mixed LHAs”)

3. Living in LHAs where no FN communities had access to prepositioned Tamiflu (“Non-Prepositioned LHAs”) This included LHAs that had no FN communities.

– Other Residents (OR)4. Living in Prepositioned LHAs5. Living in Mixed LHAs6. Living in Non-Prepositioned LHAs

Page 6: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

6

Methods: Data Sources

• Communities with prepositioned Tamiflu were identified from records & communication with FNIHB staff

• Geographical location was determined based on postal code within MSP Registration & Premium Billing File & used to assign each individual to their LHA of residence

Page 7: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

7

Methods: Data Sources

• MSP Registration & Premium Billing File: identification of cohort, age, gender, geographical location

• Status Verification File: Identification of Status Indians as of 2006

• PharmaNet: Tamiflu prescriptions filled in community pharmacies

• MSP & Discharge Abstract Database (DAD): ICD-9 & ICD-10 codes used to identify pneumonia & influenza, comorbidities1 & pregnancy

1 Using the same ICD-9 & ICD-10 codes used in national chronic disease surveillance. The case definition used was simplified because of data access limitations.

Page 8: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

8

Outcomes of Interest

Physician Billing: ICD-9 Codes Hospitalization Discharge Abstract: ICD-10 Codes

480 - Viral pneumonia481 - Pneumococcal pneumonia482 - Other bacterial pneumonia483 - Pneumonia due to other specified organism484 - Pneumonia in infectious diseases classified elsewhere 485 - Bronchopneumonia, organism unspecified 486 - Pneumonia, organism unspecified 487 - Influenza

J10 - Influenza due to other identified influenza virusJ11 - Influenza, virus not identifiedJ12 - Viral pneumonia, not elsewhere classifiedJ13 - Pneumonia due to Streptococcus pneumoniae J14 - Pneumonia due to Haemophilus influenzaeJ15 - Bacterial pneumonia, not elsewhere classifiedJ16 - Pneumonia due to other infectious organisms, not elsewhere classifiedJ17 - Pneumonia in diseases classified elsewhereJ18 - Pneumonia, organism unspecified

• Physician services & hospitalizations for influenza & pneumonia were outcomes of interest in this study

• Rates were age-standardized using the 1991 Canadian population.

• ICD-9 & ICD-10 codes were as follows:

Page 9: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

9

Results - Cohort• 49 communities had

prepositioned Tamiflu

• 6 LHAs were Prepositioned (FN pop = 8,863; 6.1% of total SI pop)

• 13 LHAs were Mixed (FN pop = 38,377; 26.2% of total SI pop)

• Remaining LHAs were Non-prepositioned (FN pop = 99,234; 67% of total SI pop)

Page 10: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

10

Results – Physician Services

Cohort

Age-standardized MSP Services

Related to Influenza Rate per 1,000 (95%

CI)

Age-standardized MSP Services Related

to Pneumonia Rate per 1,000 (95% CI)

Prepositioned FN 37.6 (33.7-42.0) 35.2 (31.0-39.86)Prepositioned Other residents 36.8 (35.6-38.1) 39.8 (38.6-41.1)Mixed FN 64.7 (62.2-67.4) 105.5 (101.6-109.5)Mixed Other residents 49.9 (49.0-50.9) 48.3 (47.5-49.2)Non prepositioned FN 70.2 (68.5-71.9) 85.0 (82.9-87.1)Non prepositioned Other residents 70.4 (70.1-70.7) 43.7 (43.5-43.9)

Table 1: Age-Standardized rate of MSP services for Influenza & Pneumonia, by cohort, January 1, 2009 –

March 31, 2010 *

*Orange cells denote a statistically significant difference in rates between FN and Other Residents in the same LHA type

Page 11: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

11

Results – Physician Services

• Prepositioned LHAs: FN had similar levels of physician services for influenza & pneumonia

• Mixed LHAs: FN had higher levels of physician services for influenza & pneumonia

• Non-prepositioned LHAs: FN had higher levels of physician services for pneumonia

Cohort

Age-standardized MSP Services

Related to Influenza Rate per 1,000 (95%

CI)

Age-standardized MSP Services Related

to Pneumonia Rate per 1,000 (95% CI)

Prepositioned FN 37.6 (33.7-42.0) 35.2 (31.0-39.86)Prepositioned Other residents 36.8 (35.6-38.1) 39.8 (38.6-41.1)Mixed FN 64.7 (62.2-67.4) 105.5 (101.6-109.5)Mixed Other residents 49.9 (49.0-50.9) 48.3 (47.5-49.2)Non prepositioned FN 70.2 (68.5-71.9) 85.0 (82.9-87.1)Non prepositioned Other residents 70.4 (70.1-70.7) 43.7 (43.5-43.9)

Page 12: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

12

Results – Physician Services

• Risk difference for FN physician visits for influenza between Non-prepositioned & Prepositioned LHAs– 70.2 - 37.6 = 32.6 or 46.4% reduction in Prepositioned LHAs

• Risk difference for FN physician visits for pneumonia between Non-prepositioned & Prepositioned LHAs– 85.0 - 35.2 = 49.8 or 58.6% reduction in Prepositioned LHAs

CohortAge-standardized MSP

Services Related to Influenza Rate per 1,000 (95% CI)

Age-standardized MSP Services Related to

Pneumonia Rate per 1,000 (95% CI)

Prepositioned FN 37.6 (33.7-42.0) 35.2 (31.0-39.86)Prepositioned Other residents 36.8 (35.6-38.1) 39.8 (38.6-41.1)Mixed FN 64.7 (62.2-67.4) 105.5 (101.6-109.5)Mixed Other residents 49.9 (49.0-50.9) 48.3 (47.5-49.2)Non prepositioned FN 70.2 (68.5-71.9) 85.0 (82.9-87.1)Non prepositioned Other residents 70.4 (70.1-70.7) 43.7 (43.5-43.9)

Page 13: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

13

Results – Hospitalizations

Table 2: Age-standardized hospitalization rate for Influenza & Pneumonia, by cohort, January 1, 2009 -

March 31, 2010

Cohort

Age-standardized Hospitalizations

Related to Influenza Rate per

1,000 (95% CI)

Age-standardized Hospitalizations

Related to Pneumonia Rate

per 1,000 (95% CI)Prepositioned FN 0.2 (0.0-1.0) 8.8 (6.7-11.3)Prepositioned Other residents 0.1 (0.1-0.2) 3.3 (3.0-3.7)Mixed FN 0.4 (0.2-0.7) 11.8 (10.5-13.2)Mixed Other residents 0.2 (0.2-0.3) 5.1 (4.8-5.4)Non prepositioned FN 0.5 (0.4-0.7) 11.3 (10.5-12.1)Non prepositioned Other residents 0.1 (0.1-0.1) 4.2 (4.1-4.2)

*Orange cells denote a statistically significant difference in rates between FN and Other Residents in the same LHA type

Page 14: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

14

Results - Hospitalizations

• FN were more likely to be hospitalized for influenza than other BC residents across the province – Only statistically significant in Non-prepositioned LHAs

• FN were more likely to be hospitalized for pneumonia than other BC residents across the province– All statistically significant, difference smallest in Prepositioned LHAs

Cohort

Age-standardized Hospitalizations

Related to Influenza Rate per

1,000 (95% CI)

Age-standardized Hospitalizations

Related to Pneumonia Rate

per 1,000 (95% CI)Prepositioned FN 0.2 (0.0-1.0) 8.8 (6.7-11.3)Prepositioned Other residents 0.1 (0.1-0.2) 3.3 (3.0-3.7)Mixed FN 0.4 (0.2-0.7) 11.8 (10.5-13.2)Mixed Other residents 0.2 (0.2-0.3) 5.1 (4.8-5.4)Non prepositioned FN 0.5 (0.4-0.7) 11.3 (10.5-12.1)Non prepositioned Other residents 0.1 (0.1-0.1) 4.2 (4.1-4.2)

Page 15: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

15

Results – Hospitalizations

• Risk difference in FN hospitalization rate for influenza between non-Prepositioned & Prepositioned LHAs– 0.5-0.2 = 0.3 or 60% reduction in Prepositioned LHAs

• Risk difference in FN hospitalization rate for pneumonia between non-Prepositioned & Prepositioned LHAs– 11.3-8.8 = 2.5 or 22% reduction in Prepositioned LHAs

Cohort

Age-standardized Hospitalizations Related to

Influenza Rate per 1,000 (95% CI)

Age-standardized Hospitalizations Related to Pneumonia Rate per 1,000

(95% CI)Prepositioned FN 0.2 (0.0-1.0) 8.8 (6.7-11.3)Prepositioned Other residents 0.1 (0.1-0.2) 3.3 (3.0-3.7)Mixed FN 0.4 (0.2-0.7) 11.8 (10.5-13.2)Mixed Other residents 0.2 (0.2-0.3) 5.1 (4.8-5.4)Non prepositioned FN 0.5 (0.4-0.7) 11.3 (10.5-12.1)Non prepositioned Other residents 0.1 (0.1-0.1) 4.2 (4.1-4.2)

Page 16: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

16

Results – Tamiflu

CohortAge-standardized

Rx Fill Rate per 1,000 (95% CI)

Prepositioned FN 24.1 (21.0-27.6) Prepositioned Other residents 22.9 (21.9-23.9)Mixed FN 34.5 (32.6-36.4)Mixed Other residents 23.2 (22.6-23.8)Non prepositioned FN 44.6 (43.2-45.9) Non prepositioned Other residents 36.2 (36.0-36.4)

Table 3: Age-standardized Prescription Dispensing Rate through PharmaNet, by cohort, January 1,

2009 – March 31, 2010

*Orange cells denote a statistically significant difference in rates between FN and Other Residents in the same LHA type

Page 17: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

17

Results - Tamiflu

• FN filled more prescriptions for Tamiflu than other residents across BC– Statistically significant in Mixed and Non-prepositioned LHAs

• Risk difference for FN Prescription fill rate between Non-prepositioned and Prepositioned LHAs:

– 44.6-24.1 = 20.5 or 46% reduction in prepositioned LHAs

• FN with comorbidities were more likely to be prescribed Tamiflu & hospitalized for pneumonia than other residents with comorbidities across BC.

CohortAge-standardized

Rx Fill Rate per 1,000 (95% CI)

Prepositioned FN 24.1 (21.0-27.6) Prepositioned Other residents 22.9 (21.9-23.9)Mixed FN 34.5 (32.6-36.4)Mixed Other residents 23.2 (22.6-23.8)Non prepositioned FN 44.6 (43.2-45.9) Non prepositioned Other residents 36.2 (36.0-36.4)

Page 18: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

18

Results - Tamiflu

• FN who received Tamiflu were more likely to be hospitalized than those that did not receive Tamiflu (OR=1.7). – Among sick individuals2, how many did or did not receive

Tamiflu3 & subsequently become hospitalized4?– Could only be conducted for Non-Prepositioned LHAs as

individuals in other LHAs may have had access to Tamiflu through their prepositioned supplies.

– This is contrary to our original hypothesis that individuals who received Tamiflu would be less likely to be hospitalized.

2 Those with an MSP claim for influenza or pneumonia. Multiple visits for influenza & pneumonia were assumed to be the same episode of illness if visits were < 2 weeks apart

3 Individuals were assumed to be taking Tamiflu if they filled a prescription in PharmaNet within ± 3 days of the physician visit

4 Within 2 weeks of their physician visit. If a prescription for Tamiflu was filled in PharmaNet < 10 days before being hospitalized, then the individual was assumed to have been hospitalized while taking Tamiflu

Page 19: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

19

Results – Rate Ratios of Outcomes

MSP Influenza

MSP Pneumonia

Hospital Influenza

Hospital Pneumonia

Rx Fill Rate

Prepositioned LHA 1.0 0.98 1.7 2.6 1.1

Mixed LHA 1.3 2.2 1.6 2.3 1.5

Non-prepositioned LHA 1.0 1.9 4.4 2.7 1.2

Table 4: Rate ratios of age-standardized rates between FN & other residents for health outcomes, by LHA type, for the Total BC

population, January 1, 2009 – March 31, 2010

Age-standardized rates were statistically higher among

FNs

Age-standardized rates were higher among FNs (but not statistically

significant)

Age-standardized rates were lower among FNs (but not statistically

significant)

• Among Prepositioned LHAs only hospitalizations for pneumonia were statistically higher for FN compared to other residents

• In Mixed LHAs, all outcomes were statistically higher among FN except for hospitalizations for influenza

• In Non-prepositioned LHAs, all outcomes were statistically higher among FN,

except for physician visits for influenza.

Page 20: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

20

Discussion

• FN generally had higher age-standardized rates of physician visits & hospitalizations for influenza & pneumonia than other BC residents

• In prepositioned LHAs, rates between FN & other residents were not statistically different for physician visits for influenza & pneumonia, hospitalization rate for influenza, or prescription fill rates for Tamiflu– This may be because of lack of access for FN to physician services and

to community pharmacies, small numbers, or because fewer individuals got sick

Page 21: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

21

Discussion

• Another study5 found that 25% of non-severe outcomes & 20.7% of ICU admissions in 13 Canadian jurisdictions were among Aboriginal people. – Our study found that 10.9% of hospital admissions for influenza

& 5.5% of hospitalizations for pneumonia were among FN, suggesting that fewer FN in BC suffered severe outcomes

– BC’s definition of ‘FN’ is different from the Aboriginal definition used in other jurisdictions & only 24% of cases in the national study answered the question on Aboriginal status.

• This lends some credence to the results of the qualitative H1N1 evaluation, in which many key informants felt that BC FN fared better in terms of fewer severe health outcomes than in other jurisdictions

5Campbell, A. R. (2010). Risk of severe outcomes among patients admitted to hospital with pandemic (H1N1) influenza. CMAJ, 182(4), 349-55.

Page 22: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

22

Limitations

• Not possible to differentiate H1N1 from other circulating subtypes of influenza

• Location of residency depended on postal codes within MSP - these may or may not be accurate

• No health care provider service data or Tamiflu dispensing data available from FNIHB nursing stations & thus the outcomes presented here are an underestimate of the effects of H1N1

• FN were identified using a 2006 file, meaning that any new registrants, including children born since 2006 would not have been included as FN in this analysis.– This would minimize the differences between Status Indian &

other resident outcomes

Page 23: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

23

Limitations

• Prophylactic Tamiflu prescriptions cannot be differentiated from Tamiflu prescribed for actual illness. This would dilute the severity outcome (OR calculations) by including prophylactic cases with actual cases

• No death data, however, very few deaths occurred• The case definitions used to identify individuals with

comorbidities & pregnant women was not as sophisticated as best-practice standards due to data access limitations

Page 24: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

24

Recommendations

• All efforts to increase access to antivirals & support the needs of remote FN communities should continue. – There is some evidence that influenza & pneumonia outcomes

were minimized among FN in LHAs with prepositioned Tamiflu

• Strategies to minimize the effects of influenza on FN in urban & semi-rural areas should be given additional consideration. – As only 6.1% of FN live in prepositioned LHAs, the greatest

potential to reduce disparities from influenza & pneumonia illness will come from additional focus on FN living in urban & semi-rural areas

• Health authorities are encouraged to work with FN & FN communities to break down silos of care & address barriers to primary care, tertiary care & community pharmacies

Page 25: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

25

Recommendations

• FN community pandemic plans should be updated on a regular basis in partnership with their local Health Authority

• A tripartite influenza surveillance strategy that coordinates surveillance information & reduces reporting burden on front-line staff should be developed

• All efforts to implement of the Aboriginal Administrative Data Standard should be supported

Page 26: “H1N1 by the numbers” Looking at H1N1 outcomes among First Nations using health records

26

Questions

• Evaluation Questions– Megan Misovic, [email protected]

• Pandemic Planning Questions– FNIHB, Tess Juliano, [email protected]