confidential: for review only - bmj...feb 17, 2020 · confidential: for review only 4 101 measles:...

Confidential: For Review O

nly

Healthy vaccinee bias at least partly explains the lower risk

of infection after MMR vaccination: evidence from the

Netherlands

Journal: BMJ

Manuscript ID BMJ.2017.038014

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 20-Feb-2017

Complete List of Authors: Tielemans, Susanne; National Institute for Public Health and the Environment, Centre for Infectious Disease Control de Melker, Hester E.; The National Institute for Public Health and the Environment (RIVM), Netherlands , Hahne, Susan; RIVM, CIE Boef, Anna; National Institute for Public Health and the Environment, Bilthoven, the Netherlands, Centre for Infectious Disease Control van der Klis, Fiona; National Institute of Public Health and the Environment (RIVM), Centre for Infectious Disease Control Netherlands, Laboratory for Infectious Diseases and Screening Sanders, Elisabeth; National Institute for Public Health and the Environment, Centre for Infectious Disease Control

van der Sande, Marianne; National Institute for Public Health and the Environment, Centre for Infectious Disease Control Knol, Mirjam J.; National Institute for Public Health and the Environment, Centre for Infectious Disease Control

Keywords: Epidemiology, Cohort study, Measles mumps rubella vaccine, diphtheria tetanus pertussis vaccine, Infectious disease, Hospitalization, Bias

https://mc.manuscriptcentral.com/bmj

BMJ


nly

1

Healthy vaccinee bias at least partly explains the lower risk of infection after MMR 1

vaccination: evidence from the Netherlands 2

3

Susanne M.A.J. Tielemans, postdoctoral researcher1 4

Hester E. de Melker, senior epidemiologist and head of surveillance and epidemiology of the 5

National Immunisation Programme1 6

Susan J.M. Hahné, consultant epidemiologist1 7

Anna G.C. Boef, postdoctoral researcher1 8

Fiona R.M. van der Klis, senior scientist immunesurveillance and head of the 9

immunesurveillance department1 10

Elisabeth A.M. Sanders, professor in pediatric immunology and infectious diseases and chief 11

scientific officer1,2

12

Marianne A.B. van der Sande, associate professor and head of the centre for epidemiology 13

and surveillance1,3

14

Mirjam J. Knol, senior epidemiologist1

15

16

1 Centre for Infectious Disease Control, National Institute for Public Health and the 17

Environment, Bilthoven, the Netherlands 18

2 Department of Pediatrics, University Medical Center Utrecht, Utrecht, the Netherlands 19

3 Julius Center for Health Sciences & Primary Care, University Medical Center, Utrecht, the 20

Netherlands 21

22

Running title: Non-specific effects of vaccination 23

Corresponding author: Mirjam J. Knol 24

National Institute for Public Health and the Environment 25

PO Box 1, 3720 BA Bilthoven, the Netherlands 26

E-mail: [email protected] 27

28

Word count total: 3922 29

Word count abstract: 306 30

31

Number of tables: 3 32

Number of figures: 2 33

Page 1 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

2

Objective: We compared risks of hospital admissions for infections between children aged ≤2 34

years who received live measles-mumps rubella vaccine (MMR) and those who received an 35

inactivated vaccine against diphtheria, tetanus, pertussis, polio, and Haemophilus influenzae 36

type b (DTaP-IPV-Hib) as most recent vaccination. 37

38

Design: Nationwide, population-based, cohort study. 39

40

Setting: In the Netherlands, DTaP-IPV-Hib + pneumococcal vaccination is recommended at 41

ages 2, 3, 4 and 11 months and MMR + meningococcal serogroup C vaccination at age 14 42

months. Data from the national vaccine register were linked to hospital admission data. 43

44

Participants: 1,096,594 children born between 2005-2011 who received the first four DTaP-45

IPV-Hib vaccines. 46

47

Main outcome measures: MMR vs. DTaP-IPV-Hib-4 as most recent vaccination in relation 48

to risk of hospitalization for infection. Cox regression was performed with most recent 49

vaccination as time-dependent variable, adjusted for sex, chronic diseases, hospitalization for 50

any reason between age 8-9 months, birth weight, gestational age, maternal age and parity, 51

parental country of birth, and postal code. Analyses were repeated with hospitalization for 52

injuries and poisoning as a negative control outcome. As part of a sensitivity analysis, risk of 53

hospitalization for infection was compared between DTaP-IPV-Hib-4 vs. DTaP-IPV-Hib-3 as 54

most recent vaccination. 55

56

Results: Having had MMR as most recent vaccination was associated with a hazard ratio 57

(HR) of 0.62 (95% CI: 0.57 to 0.67) for hospitalization for infection and 0.84 (95% CI: 0.73 58

to 0.96) for injuries and poisoning, compared with DTaP-IPV-Hib as most recent vaccination. 59

DTaP-IPV-Hib-4 as most recent vaccination was associated with a HR of 0.69 (95% CI: 0.63 60

to 0.76) for hospitalization for infection, compared with DTaP-IPV-Hib-3 as most recent 61

vaccination. 62

63

Conclusions: Our findings suggest that healthy vaccinee bias at least partly explains the 64

observed lower risk of hospitalization with infection after MMR vaccination, and that this 65

lower risk is associated with receiving an additional vaccine, and not specifically with MMR. 66

Page 2 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

3

INTRODUCTION 67

68

Vaccines against measles, diphtheria-pertussis-tetanus (DTP) and polio have led to large 69

declines in morbidity and mortality from the diseases targeted by these vaccines1,2

. It has been 70

suggested that these vaccines may also affect morbidity and mortality from infections other 71

than those targeted by the vaccines, i.e. that they have non-specific effects3. A number of 72

studies observed beneficial non-specific effects of live attenuated vaccines (such as measles 73

and BCG) and deleterious non-specific effects of inactivated vaccines (such as DTP vaccine), 74

with in general stronger effects in girls than in boys4-10

. Moreover, the sequence of vaccines 75

may be important. It has been suggested that receiving a live-attenuated measles vaccine after 76

an inactivated DTP vaccine may be associated with lower morbidity and mortality, compared 77

with receiving a DTP vaccine after or simultaneously with a measles vaccine11-16

. Which 78

immunological mechanisms could underlie these potential non-specific effects of vaccination 79

on infectious disease susceptibility is currently unknown. Trained innate immunity, which 80

depends on epigenetic reprogramming of innate immune cells, may explain some of the non-81

specific effects17,18

. Another potential mechanism is through T-cell-mediated cross-82

reactivity18

. 83

Few studies on non-specific effects have been performed in high-income countries, 84

which have low rates of infant mortality due to infectious diseases. The public health 85

relevance of non-specific effects of vaccines in high-income countries is largely unknown. 86

Two nation-wide Danish studies reported measles-mumps-rubella (MMR) as the most recent 87

vaccination to be associated with a 16% lower risk of infectious disease related hospital 88

admissions10

and a 22% lower risk of respiratory syncytial virus hospital contacts19

, compared 89

with DTP as most recent vaccine. No differences were observed between boys and girls. 90

Most of the evidence on non-specific effects of vaccines originates from observational 91

studies, which are prone to bias20

. An example is healthy vaccinee bias (also known as healthy 92

user bias or frailty bias), which occurs when children who are more susceptible to illness are 93

vaccinated later or not at all, resulting in an overestimation of the beneficial effect of the next 94

vaccination. Recently, two systematic reviews of the potential non-specific effects of DTP 95

and measles vaccines, which were commissioned by the WHO Strategic Advisory Group of 96

Experts (SAGE), were published21

. Findings suggest that receipt of measles vaccines reduces 97

overall mortality more than would be expected through their effects on the targeted diseases, 98

while DTP vaccination may be associated with an increase in all-cause mortality21

. However, 99

since most evidence came from observational studies (DTP: ten observational studies11,12,22-28

; 100

Page 3 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

4

measles: four trials6,29-31

and 17 observational studies11-13,22,28,32-42

), it was concluded that 101

these findings should be interpreted with caution. 102

We set out to investigate the risk of infectious disease related hospital admissions 103

following receipt of the live-attenuated MMR vaccine (given simultaneously with a 104

vaccination against meningococcal disease serogroup C) versus inactivated DTP-containing 105

vaccine (also includes vaccinations against polio, Haemophilus influenzae serotype b and 106

given simultaneously with a vaccination against pneumococcal disease) as most recent 107

vaccine. In addition, we explored whether healthy vaccinee bias was present in a 108

population-based nationwide cohort study of more than 1 million Dutch children. 109

Page 4 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

5

METHODS 110

111

Study design 112

The Dutch National Immunisation Programme (NIP) was implemented in 1957 and is 113

coordinated and evaluated by the Dutch National Institute for Public Health and the 114

Environment (RIVM). Table 1 and Supplementary Table 1 show the recommended vaccines 115

included in the NIP in the first 24 months of life and the years of introduction into the NIP. 116

The DTaP-IPV-Hib vaccine, recommended at ages 2, 3, 4 and 11 months, consists of 117

vaccinations against diphtheria, tetanus, pertussis (acellular), polio, and conjugate vaccine 118

against Haemophilus influenzae serotype b (Hib) and is administered simultaneously with a 119

multivalent conjugate vaccination against pneumococcal disease (PCV). The vaccination 120

against mumps, measles and rubella is recommended at age 14 months and administered 121

simultaneously with vaccination against meningococcal disease serogroup C (MenC). 122

123

Vaccination data 124

In 2005, an electronic national immunization register ‘Præventis’ was implemented in the 125

Netherlands. All children under the age of 19 years eligible for the NIP are registered and a 126

new record with a unique client number is automatically created for each newborn or 127

immigrated child. Therefore, Præventis includes a record for each child in the Netherlands, 128

irrespective of participation in the NIP. Parents receive invitation letters automatically created 129

by Præventis to get their child(ren) vaccinated at a healthy children clinic near their homes 130

according to the NIP. Data on administered vaccinations (i.e. vaccine characteristics, dose, 131

date of administration) are entered in Præventis.43

132

133

Hospital admission data 134

The National Medical Register by Dutch Hospital Data provided data on hospital admissions 135

between 1 January 2005 and 31 December 2012. Dutch Hospital Data requests hospitals and 136

university medical centres in the Netherlands to voluntarily supply hospitalization data. The 137

coverage of all hospital admissions decreased with time, with 3% missing in 2005 to 25% in 138

201244

. 139

Primary and secondary discharge diagnoses and date of hospitalization were available 140

from the register. Diagnoses were coded according to the International Classification of 141

Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). We included upper respiratory 142

infections, lower respiratory infections, gastrointestinal infections and other infections 143

Page 5 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

6

(Supplementary Table 2). Only hospital admissions that lasted >1 day (and thus included an 144

overnight stay) were included in the present study, to exclude day admissions related to 145

planned examinations and surgeries. 146

147

Covariates 148

Præventis provided data on sex, parental country of birth, and postal code, and Statistics 149

Netherlands provided data on death, migration, and parental educational level. Parental 150

educational level was classified as low (elementary education or pre-vocational education), 151

medium (senior general secondary education, pre-university education, or vocational 152

education), or high (college or university). The Netherlands Perinatal Registry provided data 153

on birth weight, gestational age, maternal age and parity (as a proxy for number of siblings). 154

155

Linkage of data sources 156

Of all 1,357,461 children that were included in Præventis, 1,356,926 (99.96%) were 157

successfully linked with the population register (the Municipal Personal Records Database) 158

using a unique personal identifier, the citizen service number. Data from the National Medical 159

Registration, Statistics Netherlands, and The Netherlands Perinatal Registry were linked to the 160

population register using probability matching, based on sex, date of birth and postal code. 161

For the National Medical Registration, unique records were successfully linked ranging from 162

97.3% in 2005 to 98.7% in 201245

. 163

164

Population for analysis 165

For the present study, vaccination data from 1,356,926 Dutch children born between 1 166

January 2005 and 31 December 2011 who were all eligible for the routine NIP were available 167

from the electronic national immunization register Præventis. Of these children, 93% received 168

the first four recommended DTaP-IPV-Hib+PCV vaccines and were eligible for inclusion in 169

this study. Children who did not receive these DTP-containing vaccinations (n=102,422) were 170

excluded to limit the possibility of bias attributable to factors related to low-vaccination 171

coverage (e.g. refusal based on religion). Moreover, children who received the fourth 172

DTaP-IPV-Hib+PCV vaccine (subsequently referred to as DTP-4) either before age 9 months 173

(n=322) or after age 20 months (n=5,235), after 31 December 2012 (n=2586), after the MMR 174

vaccine (n=18,110), or simultaneously with the MMR vaccine (n=4,333) were also excluded. 175

In addition, children who received an MMR vaccine before age 12 months (n=44,965) were 176

excluded. After excluding children with a birth weight <500 grams (n=31), with missing data 177

Page 6 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

7

on covariates (birth weight, gestational age, maternal age, parental country of birth, and postal 178

code) (n=81,389), or with missing data due to migration (n=939), 1,096,594 children (81%) 179

remained for analysis. 180

181

Statistical analysis 182

183

Main analysis 184

Cox Proportional Hazards models were used to estimate hazard ratios for infectious disease 185

related hospital admissions according to the most recent vaccination (MMR vs. DTP-4), with 186

last-received vaccination included as a time-varying variable changing at the age of receiving 187

the MMR vaccine. Children entered the model at the age of receiving DTP-4 and were 188

followed until the age of hospitalization due to infection (in case of an event) or were 189

censored at the age of death, age 24 months, age of emigration, or age on 31 December 2012, 190

whichever came first. Age was used as timescale for the Cox regression. 191

Analyses were stratified by date of birth to fully control for age, season and calendar 192

year. Associations were adjusted for sex, chronic diseases before baseline (Y/N, list of ICD-9 193

codes shown in Supplementary Table 3), hospitalization for any reason in the month before 194

baseline (DTP-4 is received from age 9 months onwards, a fixed month of 8 to 9 months was 195

chosen for this variable) (Y/N), birth weight, gestational age, maternal age and parity (1, 2, 3 196

and ≥4), parental country of birth (two Dutch parents, one Dutch and one non-Dutch parent, 197

and two non-Dutch parents) and postal code in 3 digits (as a proxy for population density). 198

Analyses were stratified by type of infection, i.e. upper respiratory, lower respiratory, 199

gastrointestinal, and other infections. Sex and birth year were examined as potential effect 200

modifiers, by an interaction test and stratified analysis. Due to many missings for parental 201

educational level, associations were additionally adjusted for parental educational level (low, 202

medium, or high) in a subsample of 79% of eligible children with complete information. 203

The same analysis was performed with hospitalization due to injury or poisoning 204

(ICD-9-CM codes 800 to 999) as a negative control outcome46

. 205

Analyses were performed using SAS version 9.2 (SAS Institute, Inc.) and Stata 14 206

(StataCorp). 207

208

Sensitivity analyses 209

First, the same analysis as described above was performed with all hospital admissions for 210

infections taken into account, thus including also day admissions without an overnight stay. 211

Page 7 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

8

Second, another analysis was performed with taking into account repeated hospital 212

admissions by using the Andersen-Gill model47

as an extension of the Cox model. All hospital 213

admissions for infection were excluded that occurred less than 14 days after a previous 214

discharge, because multiple admissions within a short period could be attributable to the same 215

infection. In case of a hospital admission, children re-entered the study 14 days after 216

discharge and were followed until a next admission or censored at the age of death, age 24 217

months or age at 31 December 2012, whichever came first. 218

Third, the same analysis as the main analysis was performed for a comparison of 219

DTP-4 versus DTP-3. Vaccination was included as a time-varying variable changing at the 220

age of DTP-4. Children entered the model from the age of receiving DTP-3 and were 221

followed until the age of hospitalization due to infection or were censored at the age of death, 222

age of MMR vaccination, age 14 months or age on 31 December 2012, whichever came first. 223

224

Patient involvement 225

No patients were involved in setting the research question or the outcome measures, nor were 226

they involved in developing plans for recruitment, design, or implementation of the study. No 227

patients were asked to advise on interpretation or writing up of results. There are no plans to 228

disseminate the results of the research to study participants or the relevant patient community. 229

Page 8 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

9

RESULTS 230

231

Study population 232

Table 2 shows characteristics of the 1,096,594 children included in the analyses. Children 233

received the DTP-4 vaccination on average (± SD) at age 11.4 ± 0.7 months and the MMR 234

vaccination on average 3.2 ± 1.0 months later. Almost all children received the MMR 235

vaccination (99.6%). The majority of children had at least one Dutch parent (89.1%) and 236

about half of the children had at least one parent with higher education (51.6% of all children 237

with data on parental educational level). About 1% of all children were hospitalized for any 238

reason between age 8-9 months and 2.5% were previously diagnosed with a chronic disease 239

before receiving the DTP-4 vaccine. 240

241

MMR vs. DTP-4 in relation to hospitalization for infection (main analysis) 242

During 1,061,242 person-years of follow-up, 10,961 children were hospitalized for >1 day 243

(26% of all hospital admissions for infection) due to an infection (admission rate 10 per 1,000 244

person-years). Admission rates declined with age, from 15 per 1,000 person-years at age 12 245

months to 7 per 1,000 person-years at age 24 months (Figure 2). Similarly, admission rates 246

declined with age in those having received MMR as most recent vaccination (average 247

admission rate of 9 per 1,000 person-years). In those with DTP-4 as most recent vaccination, 248

admission rates slightly declined until 14 months, which is the median age of receiving MMR, 249

and increased thereafter until age 17 months. Admission rates at age 17 months were 1.8-250

times higher in those with DTP-4 as most recent vaccine, compared with the overall 251

admission rates (20 vs. 11 per 1,000 person-years). The average admission rate in those with 252

DTP-4 as most recent vaccination was 14 per 1,000 person-years. 253

Having received MMR as most recent vaccination was associated with a crude hazard 254

ratio (HR) of 0.60 (95% CI: 0.55 to 0.65) for infectious disease related hospital admissions, 255

compared with DTP-4 as most recent vaccination (Table 3). After adjusting for sex, chronic 256

disease, hospitalized between age 8 and 9 months, birth weight, gestational age, maternal age 257

and parity, parental country of birth, and postal code, a HR of 0.62 (95% CI: 0.57 to 0.67) was 258

found. No effect modification was observed for sex (p for interaction=0.56) and birth cohort 259

(all p for interaction>0.15). Additional adjustment for parental educational level in the 260

subsample of children with data on parental educational level changed the results marginally 261

(HR 0.64, 95% CI: 0.58 to 0.70). Taking into account repeated hospital admissions for 262

Page 9 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

10

infections (event rate of 11 per 1,000 person-years) slightly increased the HR (0.72, 95% CI: 263

0.66 to 0.79). 264

265

Type of infection 266

Of all 10,961 infection-related hospital admissions, 43% were due to gastrointestinal 267

infections, 40% due to upper respiratory infections, 31% due to lower respiratory infections, 268

and 17% due to other infections. The HR for hospitalization according to most recent 269

vaccination ranged between 0.54 (95% CI: 0.48 to 0.62) for upper respiratory infections to 270

0.70 (95% CI: 0.61 to 0.80) for gastrointestinal infections (Supplementary Table 4). 271

272

All admissions (including day admissions) 273

When all hospital admissions for infections, including day admissions without an overnight 274

stay, were taken into account, 41,976 children were admitted to the hospital for infection 275

during 1,047,465 person-years of follow-up. Of these admissions, 83% were due to upper 276

respiratory infections, 9% due to lower respiratory infections, 12% due to gastrointestinal 277

infections, and 6% due to other infections. Having received MMR as most recent vaccination 278

instead of DTP-4 was associated with a fully-adjusted HR of 0.40 (95% CI: 0.38 to 0.41) for 279

infectious disease related hospital admissions, which was mainly driven by upper respiratory 280

infections (Supplementary Table 5). 281

282

Negative control outcome 283

During 1,068,414 person-years of follow-up, 5,150 children were hospitalized for >1 day due 284

to injury or poisoning (negative control outcome). After full adjustment, we observed a HR of 285

0.86 (95% CI: 0.75 to 0.98) for hospitalization due to injuries or poisoning for MMR as most 286

recent vaccination, compared with DTP-4 as most recent vaccination (Table 3). Also taking 287

into account admissions without an overnight stay resulted in a fully-adjusted HR of 0.80 288

(95% CI: 0.71 to 0.89) (Supplementary Table 5). 289

290

DTP-4 vs. DTP-3 in relation to hospitalization for infection 291

During 870,485 person-years of follow-up, 13,839 children were hospitalized for >1 day for 292

infection (admission rate 16 per 1,000 person-years). The average admission rate in those with 293

DTP-4 as most recent vaccination was 14 per 1,000 person-years. In those with DTP-3 as 294

most recent vaccination, admission rates were quite stable until age 11 months, which is the 295

median age of receiving DTP-4, and increased thereafter until age 14 months (Figure 2). 296

Page 10 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

11

Admission rates at age 14 months were 1.5-times higher in those with DTP-3 as most recent 297

vaccination, compared with the overall admission rates (19 vs. 13 per 1,000 person-years). 298

The average admission rate in those with DTP-3 as most recent vaccination was 17 per 1,000 299

person-years. 300

Having received DTP-4 as most recent vaccination was associated with a crude HR of 301

0.66 (95% CI: 0.60 to 0.72) for infectious disease related hospital admissions, compared with 302

DTP-3 as most recent vaccination (Table 3). After adjusting for sex, chronic disease, birth 303

weight, gestational age, maternal age, parity, parental country of birth, and postal code, a HR 304

of 0.69 (95% CI: 0.63 to 0.76) was found. Taking into account admissions without an 305

overnight stay resulted in a fully-adjusted HR of 0.48 (95% CI: 0.46 to 0.51) (Supplementary 306

Table 5). 307

Page 11 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

12

DISCUSSION 308

309

Findings of the present large-scale observational study on non-specific effects emphasize the 310

extreme difficulty in interpreting such results given the likely presence of healthy vaccinee 311

bias. In more than 1 million Dutch children aged 11 to 24 months, a 38% lower risk of 312

infectious disease related hospital admissions was found in those having had MMR+MenC as 313

the most recent vaccination, compared with those having had DTaP-IPV-Hib+PCV as the 314

most recent vaccination. We also observed a 14% lower risk of hospitalization due to injuries 315

and poisoning (negative control outcome) in those with MMR+MenC as the most recent 316

vaccination. Moreover, we observed a 31% lower risk of infectious disease related hospital 317

admissions for having the fourth DTaP-IPV-Hib+PCV as most recent vaccination, compared 318

with having the third DTaP-IPV-Hib+PCV as most recent vaccine. These findings suggest 319

that a lower risk of hospitalization is associated with adherence to the routinely recommended 320

schedule. It is likely that healthy vaccinee bias was present and (at least partly) explains the 321

lower risk of infections that we observed for receiving an additional vaccine, rather than being 322

an effect of specifically receiving MMR. 323

Our findings are in line with the conclusion of the recently published WHO-SAGE 324

review21

. We observed that after the median age of receiving the next vaccine, which was 325

MMR+MenC at 14 months in the MMR+MenC vs. DTaP-IPV-Hib+PCV analysis and the 326

fourth DTaP-IPV-Hib+PCV at 11 months in the fourth DTaP-IPV-Hib+PCV vs. third DTaP-327

IPV-Hib+PCV analysis, admission rates among those who deviated from the recommended 328

schedule suddenly increased compared to the overall admissions rates. This suggests that 329

vaccination is postponed in children that are more prone to hospitalization. An unmeasured 330

confounder could be acute illness, which may be associated with timing of vaccination and 331

risk of hospitalization. The lower risk of hospitalization due to infection for the fourth DTaP-332

IPV-Hib+PCV as most recent vaccination compared with the third DTaP-IPV-Hib+PCV, also 333

suggests that receiving an additional vaccination (and therefore adhering to the routinely 334

recommended schedule) is followed by a lower risk of hospitalization, and this is thus not a 335

finding that can be attributed to MMR specifically. This raises concerns for past and future 336

observational studies on non-specific effects and emphasizes that evidence from randomized 337

trials that also investigate different sequences of vaccines is needed to draw conclusions on 338

this matter. 339

In a nationwide population-based cohort of ~500,000 Danish children10

, the risk of 340

infectious disease related hospital admissions was 14% lower in those with MMR as their 341

Page 12 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

13

most recent vaccination compared with DTaP-IPV-Hib as most recent vaccination. There is 342

more variation in the age of receiving the MMR vaccination in Denmark than in the 343

Netherlands (median [p25 – p75]: 15.8 [15.2 – 17.0] vs. 14.3 [14.0 – 14.8]), which may be 344

attributed to a different vaccination system. In the Netherlands, appointments for each 345

vaccination are made in child health clinics in advance. In Denmark, parents have to make an 346

appointment with the GP themselves for their children to receive vaccinations. This might 347

have led to more random variation in the age of receiving the MMR vaccination and therefore 348

proportionally less variation due to factors related to the child’s health in Denmark, compared 349

with the Netherlands. The system in the Netherlands with a more fixed schedule as result of 350

prescheduled appointments more evidently illustrates the existence of healthy vaccinee bias. 351

This may explain the discrepancy between findings. However, it seems likely that healthy 352

vaccinee bias is also present in the Danish study. We consider it therefore likely that the lower 353

risk of infection as estimated in the Danish situation overestimates any non-specific effects. 354

Although the impact of potential healthy vaccinee bias could not be quantified, our 355

findings suggest that if non-specific effects of vaccines exist, that these effects would be very 356

small. This is illustrated by the minor differences in infection risk with overlapping 357

confidence intervals between the fourth vs. the third DTaP-IPV-Hib+PCV vaccinations (HR 358

of 0.69 [95% CI: 0.63 to 0.76]) and MMR vs. DTaP-IPV-Hib+PCV analyses (HR of 0.62 359

[95% CI: 0.57 to 0.67]). We consider it likely that the lower risk of infection following the 360

fourth vs. the third DTaP-IPV-Hib+PCV vaccination is due to healthy vaccinee bias, rather 361

than due to non-specific effects. These findings emphasize the difficulty of investigating non-362

specific effects of vaccination in observational studies. This is also confirmed by our findings 363

for the negative control outcome. We observed a 14% (95% CI: 4% to 27%) lower risk of 364

hospitalization due to injuries and poisoning, which can obviously not be explained by 365

non-specific effects. 366

It should be noted that in the Netherlands, the DTaP-IPV-Hib vaccine is administered 367

simultaneously with a multivalent conjugate vaccination against pneumococcal disease and 368

the MMR vaccine is administered simultaneously with vaccination against MenC. It cannot be 369

excluded that beneficial non-specific effects of the live MMR vaccine were masked by the 370

simultaneous co-administration of the attenuated MenC vaccine. 371

A major strength of our study is that it has a nation-wide population-based character 372

and included more than 1 million children. Because nearly all children received both the 373

DTaP-IPV-Hib+PCV and MMR+MenC vaccinations, children acted as their own controls. 374

Analyses were stratified by date of birth to fully control for age, year and season. The large 375

Page 13 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

14

sample size enabled us to perform multiple stratified analyses. Moreover, the Dutch 376

vaccination system, which is characterized by its fixed schedule as a result of prescheduled 377

appointments, provided the opportunity to explore and illustrate the presence of healthy 378

vaccinee bias. A limitation of the study is that we did not have information on the reason of 379

delay of vaccination. Moreover, it should be noted that not all hospital admissions were 380

captured in the National Medication Registration and that the completeness decreased with 381

time (97% in 2005 to 75% in 2012)44

. We assumed that the completeness was not associated 382

with the timing of vaccination in our study and therefore would not confound the results. The 383

interaction test confirmed this; no effect modification for birth cohort was observed. Since 384

these findings were based on hospitalization data, only severe cases were taken into account. 385

Therefore, it is of great interest to investigate the association between most recent vaccination 386

and GP consultations for infections. As this will enable us to take into account acute illness 387

(e.g. a GP visit for fever, as vaccination should be avoided if a child’s temperature is 388

≥38.5°C48

) as a time-varying variable, it may provide more insight in the effect of acute 389

illness as part of healthy vaccinee bias can have. 390

In conclusion, in this observational study on non-specific effects of vaccination in 391

more than 1 million Dutch children, a lower risk of hospitalization due to infection followed 392

receipt of an additional vaccination (and therefore adherence to the routinely recommended 393

schedule) and was not attributable to MMR specifically. Receiving MMR+MenC as most 394

recent vaccination was associated with a 38% lower risk of infectious disease related hospital 395

admissions, compared with DTaP-IPV-Hib+PCV as most recent vaccination. However, we 396

also observed a 31% lower risk of hospitalization due to infection when receiving the fourth 397

DTaP-IPV-Hib+PCV as most recent vaccination, compared with the third DTaP-IPV-398

Hib+PCV as most recent vaccination. These findings, together with those for the negative 399

control outcome, suggest that the delay of a vaccination (be it DTaP-IPV-Hib+PCV or 400

MMR+MenC) might depend on the health status of a child, rather than the other way around. 401

Even though we could not quantify the size of bias, our findings emphasize the importance of 402

interpreting findings from observational studies on non-specific effects of vaccination with 403

great caution. 404

405

Page 14 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

15

What is already known on this topic 406

- Live attenuated vaccines (such as measles) may have beneficial non-specific effects 407

and inactivated vaccines (such as DTP) may have deleterious non-specific effects. 408

- The evidence for non-specific effects of vaccines remains weak since most evidence 409

comes from observational studies that are vulnerable to confounding. 410

411

What this study adds 412

- In more than 1 million Dutch children, receiving MMR+MenC as most recent 413

vaccination was associated with a 38% lower risk of infectious disease related hospital 414

admissions, compared with DTaP-IPV-Hib+PCV as most recent vaccination. 415

- Moreover, receiving the fourth DTaP-IPV-Hib+PCV as most recent vaccination was 416

associated with a 31% lower risk of hospitalization due to infection, compared with 417

the third DTaP-IPV-Hib+PCV as most recent vaccination. 418

- These findings suggest that healthy vaccinee bias at least partly explains the observed 419

lower risk of hospitalization with infection after MMR vaccination, and that this lower 420

risk is associated with receiving an additional vaccine, and not specifically with 421

MMR. 422

Page 15 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

16

FOOTNOTES 423

424

Contributors: ST, HdM, SH, FvdK, ES, MvdS and MK contributed to the conception and 425

design of the study. ST and MK acquired and interpreted the data. ST and AB carried out the 426

statistical analysis. ST drafted the manuscript. All authors critically revised it for important 427

intellectual content and approved the final manuscript. ST and MK are the study guarantors 428

and are responsible for the decision to submit for publication. 429

430

Funding: This research was funded by the Dutch Ministry of Health, and was carried out in 431

the framework of RIVM Strategic Programme (SPR), in which expertise and innovative 432

projects prepare RIVM to respond to future issues in health and sustainability. 433

434

Competing interests: All authors have completed the ICMJE uniform disclosure form at 435

www.icmje.org/coi_disclosure.pdf and declare: support from the Dutch Ministry of Health for 436

the submitted work; no financial relationships with any organisations that might have an 437

interest in the submitted work in the previous three years; no other relationships or activities 438

that could appear to have influenced the submitted work. 439

440

Transparency: The corresponding author (MK) affirms that the manuscript is an honest, 441

accurate, and transparent account of the study being reported, no important aspects of the 442

study have been omitted, and any discrepancies from the study as planned have been 443

explained. 444

Page 16 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

17

REFERENCES 445

446

1. van Wijhe M, McDonald SA, de Melker HE, et al. Effect of vaccination programmes on 447

mortality burden among children and young adults in the Netherlands during the 20th 448

century: a historical analysis. Lancet Infect Dis 2016;16(5):592-8. doi: 449

10.1016/S1473-3099(16)00027-X 450

2. Roush SW, Murphy TV, Vaccine-Preventable Disease Table Working G. Historical 451

comparisons of morbidity and mortality for vaccine-preventable diseases in the United 452

States. JAMA 2007;298(18):2155-63. doi: 10.1001/jama.298.18.2155 453

3. Flanagan KL, van Crevel R, Curtis N, et al. Heterologous ("nonspecific") and sex-454

differential effects of vaccines: epidemiology, clinical trials, and emerging 455

immunologic mechanisms. Clin Infect Dis 2013;57(2):283-9. doi: 10.1093/cid/cit209 456

4. Kristensen I, Aaby P, Jensen H. Routine vaccinations and child survival: follow up study in 457

Guinea-Bissau, West Africa. BMJ 2000;321(7274):1435-8. 458

5. Aaby P, Samb B, Simondon F, et al. Non-specific beneficial effect of measles 459

immunisation: analysis of mortality studies from developing countries. BMJ 460

1995;311(7003):481-5. 461

6. Aaby P, Martins CL, Garly ML, et al. Non-specific effects of standard measles vaccine at 462

4.5 and 9 months of age on childhood mortality: randomised controlled trial. BMJ 463

2010;341:c6495. doi: 10.1136/bmj.c6495 464

7. Aaby P, Benn C, Nielsen J, et al. Testing the hypothesis that diphtheria-tetanus-pertussis 465

vaccine has negative non-specific and sex-differential effects on child survival in high-466

mortality countries. BMJ Open 2012;2(3) doi: 10.1136/bmjopen-2011-000707 467

8. Shann F, Nohynek H, Scott JA, et al. Randomized trials to study the nonspecific effects of 468

vaccines in children in low-income countries. Pediatr Infect Dis J 2010;29(5):457-61. 469

doi: 10.1097/INF.0b013e3181c91361 470

9. Shann F. The non-specific effects of vaccines. Arch Dis Child 2010;95(9):662-7. doi: 471

10.1136/adc.2009.157537 472

10. Sorup S, Benn CS, Poulsen A, et al. Live vaccine against measles, mumps, and rubella 473

and the risk of hospital admissions for nontargeted infections. JAMA 2014;311(8):826-474

35. doi: 10.1001/jama.2014.470 475

11. Aaby P, Nielsen J, Benn CS, et al. Sex-differential and non-specific effects of routine 476

vaccinations in a rural area with low vaccination coverage: an observational study 477

Page 17 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

18

from Senegal. Trans R Soc Trop Med Hyg 2015;109(1):77-84. doi: 478

10.1093/trstmh/tru186 479

12. Aaby P, Vessari H, Nielsen J, et al. Sex differential effects of routine immunizations and 480

childhood survival in rural Malawi. Pediatr Infect Dis J 2006;25(8):721-7. doi: 481

10.1097/01.inf.0000227829.64686.ae 482

13. Hirve S, Bavdekar A, Juvekar S, et al. Non-specific and sex-differential effects of 483

vaccinations on child survival in rural western India. Vaccine 2012;30(50):7300-8. 484

doi: 10.1016/j.vaccine.2012.09.035 485

14. Fisker AB, Ravn H, Rodrigues A, et al. Co-administration of live measles and yellow 486

fever vaccines and inactivated pentavalent vaccines is associated with increased 487

mortality compared with measles and yellow fever vaccines only. An observational 488

study from Guinea-Bissau. Vaccine 2014;32(5):598-605. doi: 489

10.1016/j.vaccine.2013.11.074 490

15. Aaby P, Biai S, Veirum JE, et al. DTP with or after measles vaccination is associated with 491

increased in-hospital mortality in Guinea-Bissau. Vaccine 2007;25(7):1265-9. doi: 492

10.1016/j.vaccine.2006.10.007 493

16. Sorup S, Benn CS, Poulsen A, et al. Simultaneous vaccination with MMR and DTaP-IPV-494

Hib and rate of hospital admissions with any infections: A nationwide register based 495

cohort study. Vaccine 2016;34(50):6172-80. doi: 10.1016/j.vaccine.2016.11.005 496

17. Blok BA, Arts RJ, van Crevel R, et al. Trained innate immunity as underlying mechanism 497

for the long-term, nonspecific effects of vaccines. J Leukoc Biol 2015;98(3):347-56. 498

doi: 10.1189/jlb.5RI0315-096R 499

18. Goodridge HS, Ahmed SS, Curtis N, et al. Harnessing the beneficial heterologous effects 500

of vaccination. Nat Rev Immunol 2016;16(6):392-400. doi: 10.1038/nri.2016.43 501

19. Sorup S, Benn CS, Stensballe LG, et al. Measles-mumps-rubella vaccination and 502

respiratory syncytial virus-associated hospital contact. Vaccine 2015;33(1):237-45. 503

doi: 10.1016/j.vaccine.2014.07.110 504

20. Farrington CP, Firth MJ, Moulton LH, et al. Epidemiological studies of the non-specific 505

effects of vaccines: II--methodological issues in the design and analysis of cohort 506

studies. Trop Med Int Health 2009;14(9):977-85. doi: 10.1111/j.1365-507

3156.2009.02302.x 508

21. Higgins JP, Soares-Weiser K, Lopez-Lopez JA, et al. Association of BCG, DTP, and 509

measles containing vaccines with childhood mortality: systematic review. BMJ 510

2016;355:i5170. doi: 10.1136/bmj.i5170 511

Page 18 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

19

22. Lehmann D, Vail J, Firth MJ, et al. Benefits of routine immunizations on childhood 512

survival in Tari, Southern Highlands Province, Papua New Guinea. Int J Epidemiol 513

2005;34(1):138-48. doi: 10.1093/ije/dyh262 514

23. Moulton LH, Rahmathullah L, Halsey NA, et al. Evaluation of non-specific effects of 515

infant immunizations on early infant mortality in a southern Indian population. Trop 516

Med Int Health 2005;10(10):947-55. doi: 10.1111/j.1365-3156.2005.01434.x 517

24. Aaby P, Ravn H, Roth A, et al. Early diphtheria-tetanus-pertussis vaccination associated 518

with higher female mortality and no difference in male mortality in a cohort of low 519

birthweight children: an observational study within a randomised trial. Arch Dis Child 520

2012;97(8):685-91. doi: 10.1136/archdischild-2011-300646 521

25. Aaby P, Jensen H, Garly ML, et al. Routine vaccinations and child survival in a war 522

situation with high mortality: effect of gender. Vaccine 2002;21(1-2):15-20. 523

26. Aaby P, Jensen H, Gomes J, et al. The introduction of diphtheria-tetanus-pertussis vaccine 524

and child mortality in rural Guinea-Bissau: an observational study. Int J Epidemiol 525

2004;33(2):374-80. doi: 10.1093/ije/dyh005 526

27. Vaugelade J, Pinchinat S, Guiella G, et al. Non-specific effects of vaccination on child 527

survival: prospective cohort study in Burkina Faso. BMJ 2004;329(7478):1309. doi: 528

10.1136/bmj.38261.496366.82 529

28. Velema JP, Alihonou EM, Gandaho T, et al. Childhood mortality among users and non-530

users of primary health care in a rural west African community. Int J Epidemiol 531

1991;20(2):474-9. 532

29. Hartfield J, Morley D. Efficacy of measles vaccine. J Hyg (Lond) 1963;61:143-7. 533

30. Benn CS, Bale C, Sommerfelt H, et al. Hypothesis: Vitamin A supplementation and 534

childhood mortality: amplification of the non-specific effects of vaccines? Int J 535

Epidemiol 2003;32(5):822-8. 536

31. Aaby P, Garly ML, Nielsen J, et al. Increased female-male mortality ratio associated with 537

inactivated polio and diphtheria-tetanus-pertussis vaccines: Observations from 538

vaccination trials in Guinea-Bissau. Pediatr Infect Dis J 2007;26(3):247-52. 539

32. Van Balen H, Mercenier P, Daveloose P. Influence of measles vaccination on survival 540

pattern of 7--35-month-old children in Kasongo, Zaire. Lancet 1981;1(8223):764-7. 541

33. Aaby P, Samb B, Simondon F, et al. Divergent mortality for male and female recipients of 542

low-titer and high-titer measles vaccines in rural Senegal. Am J Epidemiol 543

1993;138(9):746-55. 544

Page 19 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

20

34. Kabir Z, Long J, Reddaiah VP, et al. Non-specific effect of measles vaccination on overall 545

child mortality in an area of rural India with high vaccination coverage: a population-546

based case-control study. Bull World Health Organ 2003;81(4):244-50. 547

35. George K, Joseph A, Muliyil J, et al. Measles vaccination before nine months. Trop Med 548

Int Health 1998;3(9):751-6. 549

36. Holt EA, Boulos R, Halsey NA, et al. Childhood survival in Haiti: protective effect of 550

measles vaccination. Pediatrics 1990;85(2):188-94. 551

37. Fisker AB, Hornshoj L, Rodrigues A, et al. Effects of the introduction of new vaccines in 552

Guinea-Bissau on vaccine coverage, vaccine timeliness, and child survival: an 553

observational study. Lancet Glob Health 2014;2(8):e478-87. doi: 10.1016/S2214-554

109X(14)70274-8 555

38. Aaby P, Knudsen K, Jensen TG, et al. Measles incidence, vaccine efficacy, and mortality 556

in two urban African areas with high vaccination coverage. J Infect Dis 557

1990;162(5):1043-8. 558

39. Aaby P, Bukh J, Lisse IM, et al. Measles vaccination and reduction in child mortality: a 559

community study from Guinea-Bissau. J Infect 1984;8(1):13-21. 560

40. Aaby P, Bukh J, Lisse IM, et al. Determinants of measles mortality in a rural area of 561

Guinea-Bissau: crowding, age, and malnutrition. J Trop Pediatr 1984;30(3):164-8. 562

41. Aaby P, Bhuiya A, Nahar L, et al. The survival benefit of measles immunization may not 563

be explained entirely by the prevention of measles disease: a community study from 564

rural Bangladesh. Int J Epidemiol 2003;32(1):106-16. 565

42. Breiman RF, Streatfield PK, Phelan M, et al. Effect of infant immunisation on childhood 566

mortality in rural Bangladesh: analysis of health and demographic surveillance data. 567

Lancet 2004;364(9452):2204-11. doi: 10.1016/S0140-6736(04)17593-4 568

43. van Lier A, Oomen P, de Hoogh P, et al. Praeventis, the immunisation register of the 569

Netherlands: a tool to evaluate the National Immunisation Programme. Euro Surveill 570

2012;17(17) 571

44. Documentatierapporten LMR (Documentation reports National Medical Register): 572

Centraal Bureau voor de Statistiek (Statistics Netherlands), 2005 - 2012. 573

45. Bruin de A, Ariel A, Verweij G, et al. Methode van bijschatten van StatLinetabel 574

Ziekenhuispatiënten naar diagnose. Den Haag/Heerlen: Centraal Bureau voor de 575

Statistiek (Statistics Netherlands), 2009. 576

Page 20 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

21

46. Lipsitch M, Tchetgen Tchetgen E, Cohen T. Negative controls: a tool for detecting 577

confounding and bias in observational studies. Epidemiology 2010;21(3):383-8. doi: 578

10.1097/EDE.0b013e3181d61eeb 579

47. Andersen PK, Gill RD. Cox's Regression Model for Counting Processes: A Large Sample 580

Study. Ann Stat 1982;10(4):1100-20. 581

48. National Institute for Public Health and the Environment (RIVM). [Guidelines for 582

contraindications for vaccination]. Available from: 583

http://www.rivm.nl/Documenten_en_publicaties/Professioneel_Praktisch/Richtlijnen/I584

nfectieziekten/Rijksvaccinatieprogramma/Uitvoeringsregels_RVP_2015_2016/Inhoud585

/2_Contra_indicaties accessed January 11, 2017. 586

Page 21 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

22

TABLES 587

588

Table 1. Schedule of the NIP from 2005 to 2011 until age 24 months

Age (in months) Vaccination 1 Vaccination 2

2 DTaP-IPV-Hib(-HepB since August 2011) PCV




14 MMR MenC

Abbreviations: DTaP, diphtheria-tetanus-pertussis (acellular); HepB, hepatitis B; Hib,

Haemophilus influenzae serotype b; IPV, inactivated polio vaccine; MenC, meningococcal

disease serogroup C; MMR, measles-mumps-rubella; NIP, National Immunisation

Programme; PCV, pneumococcal conjugate vaccine.

589

590

Page 22 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

23

Table 2. Characteristics of 1,096,594 children included in the present study

Age fourth DTaP-IPV-Hib vaccination (days) 342 (336 – 354)

Age MMR vaccination (days)1

435 (427 – 451)

Male (%) 561 407 (51.2%)

Birth weight (grams) 3,460 (3,100 – 3,810)

Gestational age (weeks) 39.9 (38.7 – 40.7)

Maternal age at birth of child (years) 31 (27 – 34)

Maternal parity2

One 505,851 (46.1%)

Two 400,703 (36.5%)

Three 138,565 (12.6%)

Four or more 51,473 (4.7%)

Highest parental educational level3,4

Low 115,205 (13.2%)

Medium 310,445 (35.6%)

High 445,544 (51.1%)

Parental country of birth (%)

The Netherlands 834,580 (76.1%)

The Netherlands and foreign 142,555 (13.0%)

Foreign 119,457 (10.9%)

Chronic disease (%) 27,430 (2.5%)

Hospitalization for any reason between age 8 – 9 months (%) 11,706 (1.1%)

Values are presented as median (p25 – p75) or N (%).

1 Data are available for 1,092,625 children;

2 Represents the number of childbirths of the mother;

3 Highest educational level of the household (mother or father);

4 Data are available for 871,194 children.

591

592

Page 23 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only24

Table 3. HR (95% CI) for hospital admission due to infections (outcome of interest) and due to injuries or poisoning (negative

control outcome) according to the most recent vaccination

Outcome

Most recent vaccination

Events / sum person-years

HR (95% CI)

Crude Fully-adjusted1

All infections DTP-4

MMR

4,111 / 284,786 1.00 1.00

6,850 / 776,456 0.60 (0.55 to 0.65) 0.62 (0.57 to 0.67)

Injuries or poisoning DTP-4

MMR

3,605 / 285,676 1.00 1.00

5,150 / 782,738 0.81 (0.71 to 0.93) 0.84 (0.73 to 0.96)


DTP-4

10,654 / 639,484 1.00 1.00

3,185 / 231,001 0.66 (0.60 to 0.72) 0.69 (0.63 to 0.76)2

1 Adjusted for sex, chronic disease, hospitalized for any reason between age 8 and 9 months, birth weight, gestational age,

maternal age and parity, parental country of birth, and postal code;

2 Adjusted for sex, chronic disease, birth weight, gestational age, maternal age and parity, parental country of birth, and postal

code.

Page 24 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

25

Figure 1. Proportion of children according to most recent vaccination, with

DTaP-IPV-Hib+PCV in light grey and MMR+MenC in dark grey, based on 1,096,594

children included in the present study

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

11 12 13 14 15 16 17 18 19 20 21 22 23 24

Per

cen

tage

vac

cin

ated

(%

)

Age (in months)

Page 25 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nly

26

Figure 2. Hospital admissions rates for infection with 95% confidence intervals by most

recently received vaccination: DTP-4 vs. MMR (top) and DTP-3 vs. DTP-4 (bottom). The

dashed vertical lines represent the recommended age of vaccination (top: MMR at age 14

months; bottom: DTP-4 at age 11 months). Each estimate represents the incidence of the

month prior to that month, e.g. the estimate at age 12 months comprises the incidence

between age 335 – 364 days.

0

5

10

15

20

25

30

11 12 13 14 15 16 17 18 19 20 21 22 23 24

Inci

den

ce p

er 1

,00

0 p

erso

n-y

ears

Age (in months)

Overall

DTP-4

MMR

0

5

10

15

20

25

30

4 5 6 7 8 9 10 11 12 13 14

Inci

den

ce p

er 1

,00

0 p

erso

n-y

ears

Age (in months)

Overall

DTP-3

DTP-4

Page 26 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nlySUPPLEMENTARY TABLES

Supplementary Table 1. Introduction of diseases into the NIP in children aged ≤2 years

Year Disease

1957 Diphtheria, pertussis, tetanus, polio

1974 Rubella

1976 Measles

1987 Mumps

1993 Haemophilus influenza type b

2002 Meningococcal disease serogroup C

2006 Pneumococcal disease

2011 Hepatitis B

Abbreviation: NIP, National Immunisation Programme.

Page 27 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nlySupplementary Table 2. List of ICD-9 codes per type of infection

Type of infection ICD-9 codes

Upper respiratory 034.0, 381.0 – 381.6, 382.0 – 383.2, 383.8 – 383.9, 460 – 465,

473, 474.0, 475, 476.0

Lower respiratory 073.0, 466, 485 – 486, 480, 482 – 483, 484.1, 484.5 – 484.8,

487 – 488, 510, 511.0, 513

Gastrointestinal 001 – 009, 021.1 – 021.2, 088.0, 127, 129, 136.5, 535.0, 535.4

– 535.6

Other infections 010 – 018, 020 – 027, 030 – 031, 034.1, 035, 036.0, 036.4 -

036.9, 038.0 – 038.1, 038.3 – 038.9, 039 – 040, 041.0 – 041.1,

041.3 – 041.4, 041.6, 041.8 – 041.9, 045 – 048, 049.0 – 049.1,

049.8, 052.0, 053.0, 054.0, 054.2 – 054.9, 055.0, 056.0, 061 –

066, 070.0 – 070.1, 070.4 – 070.7, 071, 072.1 – 072.2, 076,

077.98, 078.1, 078.81, 078.3, 079.89, 079.98, 080 – 083 –

086, 087, 088.8 – 088.9 , 099.2, 100 – 104, 110 – 134, 133.8

– 133.9, 136.2 – 136.9, 320 – 322, 323.0 – 323.6, 323.8 –

323.9, 372.0, 372.2 – 372.3, 373.0 – 373.2, 373.4 – 373.9,

380.0 – 380.2, 390 – 391, 420 – 421, 424.9, 595.0, 597, 590.1

– 590.3, 590.9, 680 – 683, 686.0, 686.8 – 686.9, 690.8,

695.81, 684, 711.0, 711.4 – 711.8, 728.0, 730.0 – 730.2, 730.8

– 730.9, 995.91

Page 28 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nlySupplementary Table 3. List of ICD-9 codes for chronic diseases

Type of disease ICD-9 codes

Cancer 140 - 209

Congenital 740 - 759

Nervous system 331.3 – 331.5, 335, 343, 345, 356, 359.0 – 359.2

Endocrine, metabolic and immunity 250, 270 – 272, 275, 277.0 – 277.2, 277.4 – 277.5, 279

Circulatory system 416, 420 – 429

Genitourinary system 581 – 583, 585 – 589, 591, 593.3 – 593.7

Respiratory system 515 – 516, 770.7, 771.0 – 771.1

Digestive system 570 – 573, 530.11, 530.81

Autoimmune 131.1, 446, 710, 725

Page 29 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

Confidential: For Review Only

Supplementary Table 4. HR (95% CI) for hospital admission due to infectious disease according to the most recent

vaccination, stratified per type of infection

Type of infection

Most recent vaccination

Events / person-years

Hazard ratio (95% CI)


Upper respiratory DTP-4 1,664 / 285,664 1.00 1.00

MMR 2,765 / 782,900 0.50 (0.44 to 0.57) 0.54 (0.48 to 0.62)

Lower respiratory DTP-4 1,234 / 285,731 1.00 1.00

MMR 2,120 / 783,457 0.52 (0.44 to 0.57) 0.56 (0.49 to 0.66)

Gastrointestinal DTP-4 1,816 / 285,657 1.00 1.00

MMR 2,942 / 782,668 0.67 (0.59 to 0.76) 0.70 (0.61 to 0.80)

Other DTP-4 619 / 285,841 1.00 1.00

MMR 1,199 / 784,270 0.61 (0.50 to 0.75) 0.63 (0.52 to 0.77)

1 Adjusted for sex, chronic disease, hospitalization for any reason between age 8 and 9 months, birth weight, gestational

age, maternal age and parity, parental country of birth, and postal code.

Page 30 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


nlySupplementary Table 5. HR (95% CI) for all hospital admission (including day admissions

without an overnight stay) due to infectious disease according to the most recent vaccination

Outcome

Most recent

vaccination

Events /

person-years

Hazard ratio (95% CI)


All infections DTP-4 14,748 / 283,780 1.00 1.00

MMR 27,228 / 763,685 0.38 (0.37 to 0.40) 0.40 (0.38 to 0.41)

Upper respiratory DTP-4 12,063 / 284,226 1.00 1.00

infections MMR 22,891 / 767,395 0.34 (0.33 to 0.36) 0.36 (0.34 to 0.37)

Lower respiratory DTP-4 1,366 / 285,706 1.00 1.00


Gastrointestinal DTP-4 1,991 / 285,631 1.00 1.00


Other infections DTP-4 769 / 285,819 1.00 1.00

MMR 1,591 / 784,015 0.59 (0.49 to 0.71) 0.61 (0.51 to 0.73)

Injuries and DTP-4

MMR

1,933 / 269,677 1.00 1.00

poisoning 4,954 / 786,171 0.78 (0.69 to 0.87) 0.80 (0.71 to 0.89)


DTP-4

17,908 / 638,258 1.00 1.00

9,960 / 229,014 0.47 (0.45 to 0.49) 0.48 (0.46 to 0.51)2

1 Adjusted for sex, chronic disease, hospitalization for any reason between age 8 and 9 months,

birth weight, gestational age, maternal age and parity, parental country of birth, and postal code;

2 Adjusted for sex, chronic disease, birth weight, gestational age, maternal age and parity,

parental country of birth, and postal code.

Page 31 of 31


BMJ

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

confidential: for review only - bmj...feb 17, 2020 · confidential: for review only 4 101 measles:...

Documents