Writing Superior, Clear and Innovative Specific Aims · PDF file1 Writing Superior, Clear and Innovative Specific Aims R2 Workshop October 24, 2017 Pam Factor-Litvak

1

Writing Superior, Clear and

Innovative Specific Aims

R2 Workshop

October 24, 2017

Pam Factor-Litvak

2

www.theresearchassistant.com/tutorial

3

Preparatory Work

Answer the Following Questions

• Do I know the field and its literature well?

• What are the important research questions in my field?

• What areas need further exploration?

• Could my study fill a gap? Lead to greater understanding?

• Has a great deal of research already been conducted in this topic area?

• Has this study been done before? If so, is there room for improvement?

• Is the timing right for this question to be answered? Is it a hot topic, or is it becoming obsolete?

• Would funding sources be interested?

• If you are proposing a service program, is the target community interested?

• Most importantly, will my study have a significant impact on the field?

4

So What??

• A strong research idea should pass the “so what” test. Think about the potential impact of the research you are proposing.

– What is the benefit of answering your research question?

– Who will it help (and how)? If you cannot make a definitive statement about the purpose of your research, it is unlikely to be funded.

• A research focus should be narrow, not broad-based.

– For example, “What can be done to prevent substance abuse?” is too large a question to answer. It would be better to begin with a more focused question such as “What are the associations between specific early childhood experiences and subsequent substance-abusing behaviors?”

5

Be Aware of Your Long Term

Goals• Why are you doing this research?

• What are the long-term implications?

• What will happen after the grant?

• What other avenues are open to explore?

• What is the ultimate application or use of the

research?

6

The Cornerstone of a Research Proposal

• What are the objectives of the work?

Master Plan of your research.

Simple and easy to read.

Dense, full-of-jargon, poorly-written Specific Aims will

not help the review of your proposal, even if the

science is sound!

Is a useful summary for obtaining early feedback on

your proposal

7

The Science of Specific Aims

• Unmet scientific need

• Includes project goals, hypotheses to be without

fine detail

• Original ideas

• Novelty of design

• Fills a gap in the existing scientific story

8

The Art of Specific Aims

• Be crystal-clear in your writing!

– State the conceptual framework

– Limit scientific jargon

– Non-experts in the field should understand the aims and

why they are important

9

The Marketing of Specific Aims

• Review committee considerations

– Not all members read every word of the proposal

– Many read just the specific aims

• Statement of aims preceded by several paragraphs

– National and big picture science

– Clinical and translational importance

10

Remember Storyboarding…

Grant Preparation Storytelling Storyboard:

Each box represents one aim of your research story. Describe in as much detail as possible what will be included in each aim. The more you plan ahead of time, the easier it is to write your proposal. Copy as many storyboard boxes as you need to plan your digital story. These will be refined as you progress.

Aim (Goal) #1 (general or specific):

Significance (bullet points):

Methods to Achieve Aim (bullet points)):

11

The Formula for Specific Aims

1. Introductory Paragraph

Opening sentence containing extremely creative

thoughts. This will grab the attention of your reader.

State what is known about this issue.

State what is unknown about this issue.

State why is this lack of knowledge is important.

12

Formula, cont’d2. What, Why Whom Paragraph

State the overall, long-term goal of your mission. This represents any future projections or the

continuum of your line of research.

State the overall objective of this application. This is a step to achieving your long term goal.

Steps along the continuum for this research that must be achieved, regardless of how hypothesis

tests.

Clearly state your central hypothesis. All vague and unfocused fishing expeditions for information

will not adequately fulfill your research mission and appears as an invalid study design.

State your rationale for your hypothesis, or how did you come up with the central hypothesis.

Briefly explain why your research design are the best possible solution for the topic at hand.

13

Examples• The overarching goal of this proposal is to estimate

associations between prenatal exposure to mixtures of

correlated endocrine disrupting chemicals (EDCs) and

child neurodevelopment.

• The goal of our research is to determine the mechanism

of X (or Y, or Z).

• The overall goal of our research group is to define the

role of …

14

Mounting evidence suggests that prenatal exposure to these EDCs may be associated

with adverse child neurodevelopment, yet previous studies are limited by small sample

sizes, inconsistencies in developmental measures, and differences in the timing of

exposure and outcome assessments1,2. More importantly, previous research has been

hindered by a fundamental methodological problem: the challenge of estimating the

neurodevelopmental impact of complex mixtures of highly inter-correlated EDC

chemicals, with varying concentrations of constituents and potentially different effects

across populations. Only recently have investigators begun developing and applying

methods to study the effects of complex environmental chemical mixtures, and these

methods have yet to be applied to the study of EDC effects on neurodevelopment.

This is a critical research gap, as exposures to EDCs are ubiquitous in the US and

other developed countries. The validation of methods for evaluating effects of chemical

mixtures has been identified as a high priority by NIEHS and the National Academy of

Sciences3,4.

15

Formula, cont’d3. Specific Aims

Detail your specific aims that will test your central hypothesis,

citing primary and secondary endpoints.

Write in the ACTIVE voice.

Use verbs:

to explain

to examine

to investigate

to estimate

to compare

16

ExampleSpecifically, we propose to:

Estimate the independent and combined associations between prenatal exposure to

mixtures of correlated EDCs metabolite concentrations and age 6-7 year

neurodevelopmental endpoints in the SELMA cohort based on the results of a

comparative modeling strategy.

Compare the results of four different analytic strategies to model exposure-outcome

associations in the SELMA cohort;

Select the strategy(s) that best meets criteria of congruence between discovery and

confirmation phases and correct identification of etiologically important components

in the simulation phase.

Apply the selected analytic strategy(s) from Aim 1 to a New York City cohort

(CCCEH) to ascertain whether these methods are applicable in a heterogeneous

population with different EDC exposure concentrations.

17

Formula, cont’d4. Payoff paragraph

Briefly explain why this application is innovative

State plainly and simply the general positive impact that

your study will have on science.

Why should the reviewers care

Why you should get funded

Why is it important for the general population

18

ExampleResults will inform future researchers and policy

makers about possible developmental risks associated

with prenatal exposure to EDCs across populations of

children.

19

Hurricane Sandy Grant

20

Major climatic events, such as hurricanes, appear to be increasing due to the

consequences of global warming. Such events are likely associated with

increased psychological stress. On October 29, 2012 Superstorm Sandy, a

major hurricane, devastated the mid-Atlantic region of the United States,

particularly the New York City/New Jersey area. Along the coastline, houses

and businesses were destroyed and residents were evacuated; as of February

2013 some residents were still displaced. In Manhattan, few residents were

displaced; those who were quickly returned. However, the storm resulted in

serious perturbations to mass transportation, which led to income loss for

many. Further, the storm also resulted in long-lasting power outages and

shortages of basic supplies such as bread, milk and gasoline.

21

Pregnant women are considered a vulnerable population. There is

increasing evidence that acute psychosocial stressors may be associated

with adverse pregnancy outcomes, such as decreases in birth weight

and decreases in gestational length. Such outcomes may be the result of

acute biological responses to stressors, such as increased production of

hormones such as cortisol and increased inflammatory processes.

Previous studies in pregnant women following natural disasters find

associations with depressive symptoms, anxiety and post-traumatic

stress syndrome, and with birth weight, preterm delivery and

intrauterine growth restriction; inferences from these are limited,

however, because baseline measures of perceived stress were not

available and sample sizes were limited.

22

One recent novel hypothesis relates psychosocial stress and

experiences of adverse life events to decreases in leukocyte

telomere length (LTL). LTL is associated with chronic

diseases, particularly atherosclerotic heart disease, in

adulthood and decreased longevity. Little is known regarding

the determinants of LTL at birth, which is likely the largest

predictor of LTL in adulthood. Given the paucity of literature,

it is of interest to determine whether exposure to acute

stressors in pregnancy is associated with decrements in LTL

at birth.

23

We were fortunate to be in the process of recruiting 1000 mother-father-newborn trios for a study

of prenatal determinants of leukocyte telomere length at birth (R01 HD071180) at the time of

Superstorm Sandy. The trios come from prenatal clinics at Columbia University Medical Center in

Manhattan and Christiana Health Care Center in Delaware. The protocol for the parent study is

remarkable for: comprehensive evaluation of fetal growth by ultrasound from early in pregnancy to

parturition; comprehensive measurements of maternal lifestyle, behavior, and health; and the archiving

of biological specimens from mothers and newborns. In this R21 application we build upon this study

and propose to evaluate the effects of Superstorm Sandy on pregnancy outcomes and LTL in

newborns. Because of the timing of recruitment, we are able to evaluate several “exposure windows”,

that is no exposure (i.e. birth before the storm), exposure in each trimester and in the three months

prior to pregnancy. We also are fortunate to have an unexposed site to serve as a “place” control. For

several important reasons, the proposed study improves upon previous work on stress and pregnancy:

first, we have both time and place controls and second, we have baseline assessments of stress, social

support, anxiety, depression, and resilience. Further, we have sufficient sample size to test our

hypotheses. Finally, we are already administering a questionnaire to participants asking about their

experiences and stress during the storm.

24

The specific aims are:

1. To examine associations between exposure to Superstorm Sandy and a)

decreased length of gestation, b) reduced birth weight, c) reduced birth

weight for gestational age, d) decreased head circumference, e) decreased

birth length. Secondary aims will examine the associations between

exposure in specific trimesters and outcomes and between exposure and

dichotomous outcomes such as preterm birth and small for gestational age

(SGA).

2. To examine associations between exposure to Superstorm Sandy and LTL

in newborns and to examine whether associations, if any, are modified by

baseline levels of perceived stress, social support, anxiety, depression and

resilience.

25

Our results have the potential to inform emergency responders and

clinicians how best to support and potentially mitigate the effects of

psychological stress among pregnant women during and after a

major natural disaster. This study will also set the stage for studies

to inquire whether exposure to stressful events during the fetal

period has long lasting effects on behavior and cognition in children.

26

www.usuhs.mil/medschool/faculty/ppt/writingspecificaims.pptx

27

Finally…

– Propose good science

–Write well

–Get feedback

28

Sample Specific Aims

29

Specific Aims Example 2017 (A 1/4)

Tuberculosis (TB) and HIV/AIDS are linked in a deadly syndemic with bothclinical and behavioral components. In South Africa interactions betweenmultidrug resistant or extensively drug resistant TB (MXDR-TB) andgeneralized HIV/AIDS epidemics results in high mortality for people living withHIV (PLWH) and threatens both HIV and TB control programs. Medicationadherence is arguably the key determinant of treatment outcomes in both HIVand MXDR-TB (4, 5), and intermittent or sub-optimal adherence can generateresistant strains in each. Our group has played a substantial role in advancingour understanding of medication adherence in MXDR-TB/HIV treatment anddeveloping implementation science strategies to improve clinical outcomes.

Prospective data from our group shows low and differential adherence to TBmedications compared to ART in M/XDR-TB/HIV patients in KwaZulu-Natal,South Africa (8). Consequences may be critical and include amplification ofdrug resistance on treatment, community spread of resistant TB and HIV andpoor treatment outcomes.

30

Specific Aims Example 2017 A 2/4

Our research has also highlighted the role of stigma in shaping patient experiences of inpatient and community treatment of MXDR-TB among PLWH. Surveys suggest that over half of people co-infected with MXDR -TB and HIV feel ashamed of having TB, and 80% reported an experience of stigma due to TB and/or HIV. Numerous calls to end the TB epidemic highlight the role of stigma (6) (7-9), and highlight the effect of HIV co-infection in amplifying experiences of stigma(10). Despite a growing understanding of the importance of the impact of stigma on TB/HIV treatment, our understanding of mechanisms through which stigma compromises medication adherence, or the effect of behavioral interventions to mitigate M/XDR-TB/HIV stigma are poorly understood.

To address this gap, we propose a mixed methods study nested within a funded randomized implementation science trial of an adherence intervention for MXDR-TB/HIV treatment among patients on ART and initiating Bedaquiline (a novel second-line antimycobacterial agent) using a real-time electronic medication monitoring integrated with a model-informed, community-based, behavioral intervention to improve HIV viral load, MXDR-TB culture conversion and combined clinical outcomes.

31

Specific Aims Example 2017 A 3/4

Aim 1 To measure stigma to TB and HIV as a potential mediator of differential medication adherence to Bedaquiline and ART and predictor of biological outcomes

Hypothesis 1a. Stigma to TB and HIV measured by separate validated scales will predict medication adherence to Bedaquiline and ART measured by an real-time electronic medication monitor (Wisepill).

Hypothesis 1b. Baseline stigma will be associated with HIV viral load and TB treatment response for MXDR-TB/HIV patients initiating treatment with Bedaquline and ART.

32

Specific Aims Example 2017 A 4/4

Aim 2. To determine the effect of an adherence support intervention to modify MXDR-TB/HIV associated stigma within the context of a randomized controlled trial.

2. A comprehensive adherence support intervention will reduce MXDR-TB/HIV associated stigma as measured by validated scales in the intervention arm compared to control.

Aim 3. To identify potential differences between DR-TB and HIV stigma between in-patient and in-community cohorts. To provide evidence-informed recommendations for optimizing treatment outcomes, we will evaluate data for challenges across stigma dimensions that are more acute during DR-TB treatment in hospital settings and those that are more acute for patients continuing their TB-treatment in their community, controlling for time since diagnosis.

33

Specific Aims Sample 2017 B

To establish a clear range of costs and impacts associated with systematically utilizing evidence in public policies related to energy usage in US cities. This value statement will be used to compare between approaches for developing and implementing policies, particularly in terms of their use of scientific evidence.

To identify useful interventions for further application in improving energy choices at the community level.

To test a theoretical model for how populations respond to policies in social dilemmas, using the context of energy choice.

To utilize insights and precise outcomes from objectives 1 to 3 for application into a broader framework for effective integration of scientific evidence into policy. This framework should be useful for a number of policy domains but emphasize outcomes of economic stability, security and safety, and population well-being.

34

Specific Aims Example #3

• Being childless is associated with worse health for both men and women. In recent years, childlessness, particularly for men, is increasing in many countries. In 2014, for example, approximately one in four 45-year old men were childless, compared to one in eight men in 1990, while the proportion of men with relatively high fertility (i.e. 3 or more children) remained about the same. In this proposal we address whether fertility, defined as the number of children a man has fathered, and fertility sequences, which account for spacing between births, are risk factors for poorer health in men aged 50 years and above. While fertility problems have been public-health concerns for a long time, in times of population aging, the association between fertility and late-life health outcomes has taken on additional importance. Research suggests that reduced fertility may be a marker of poorer general health and shorter life expectancy (Eisenberg et al 2016; 2014; Hanson et al 2016; Walsh et al 2010; Latif et al 2017; Jensen et al 2009), but the body of literature is still equivocal. We posit that inconsistencies may be due to a lack of appreciation of the biopsychosocial aspects of fertility behavior such as biological preconditions, stress, selection into childbearing and cultural and familial norms regarding childrearing.

• Additional methodological problems make previous studies examining associations between male fertility, fertility sequences and health outcomes problematic. To start, many of these studies do not include men who are childless (or who are not seeking treatment for fertility problems), thus introducing potential selection bias. Further, they mostly lack sufficiently rich data and longitudinal data allowing one to establish both fertility sequences and the temporal relationships between fertility, fertility sequences and health outcomes. Even if they have such data, they lack data on social and health-related variables, thus not allowing important effect modification and mediation to be analyzed. Regarding outcomes, most studies focus on a single outcome, such as self-rated mental health, all-cause mortality or specific disease groups such as cancer or cardiovascular disease (CVD).

35

• There are several mechanisms by which fertility status may be related to later-life outcomes. Here we consider two separate but not mutually exclusive pathways, namely social circumstances such as cultural and family norms regarding fertility and childbearing, and biological variables such as the stress related to infertility or to childrearing. Further, we note fertility is dynamic and the sequence in which men bear children may differ greatly and may vary with social circumstances and biological variables. Fertility status measured at one point in time does not reflect this dynamic interplay. Hence, we propose studying fertility sequences as the ‘exposure’. The proposed study aims to assess how fertility and fertility sequences relate to health using two health outcome categories: 1) Morbidity and mortality from cardiovascular disease , and 2) DNA methylation age relative to chronological age, as a marker of accelerated or slowed biological aging (ASA). We will also consider mental health (depression) and all cause mortality as secondary outcomes.

• We are fortunate to have access to the unique data collected as part of the Cohort of Norway (CONOR) consortium. These cohorts can be merged with state-of-the-art registry data collected by the Norwegian government: the Primary Health Care Registry, which documents all primary care encounters, the Prescription Drug Registry, which documents all pharmaceuticals dispensed, and the Cause of Death Registry. Further, we have objective data on the children these men have fathered (and the spacing) from the Norwegian Birth Registry. Hence, this large-scale population based study will have minimal sample bias. Analyses will include data on risk factors (from representative surveys), disease diagnoses (from primary health care and hospital records), mortality (from cause of death registry information) and biomarkers such as DNA methylation data (from Norwegian biobanks). The specific aims are:

•

36

• 1. To examine associations between fertility status/sequences and (a) CVD and mortality over age 50 and (b) predicted DNA methylation age compared to chronological age. In secondary analyses we will consider mental health and all cause mortality.

• 2. To examine the robustness of these relationships by (a) accounting for a rich set of socioeconomic, health-related and contextual variables; (b) controlling for unobserved time-invariant individual factors (e.g. using longitudinal fixed-effects models), (c) using statistical approaches, such as instrumental variables techniques (i.e. factors causing an increase in number of children not related to the health outcomes), to overcome any potential selection bias.

• 3. To examine whether risky behaviors (smoking, drinking), social stressors, such as marriage or partnership disruptions, spousal bereavement or social and family norms modify or mediate the associations between fertility status/sequence and (a) health outcomes and (b) DNA methylation age compared to chronological age.

37

• Given the access to the Norwegian data sets, the

proposed study is in a unique position to clarify the

inconsistent findings on the association between

male fertility and health in later life. The results of

the study will provide valuable information regarding

the implications of childbearing and childrearing on

later-life health that will lend itself to educational

public-health interventions.

38

Thank you!

Research Resources

Pam Factor – Litvak, Associate Dean, Research Resources and Professor of

Epidemiology at CUMC, [email protected]

Craig Kandell, Project Coordinator, Research Resources, [email protected]