| |
|
New Approaches in H2R Research Methods | Sample size calculation for a PSE project |
Fearon et al. published a paper in JMIR 2017 to provide guidance for calculating the sample size in population size estimation (PSE) studies using multiplier methods to minimize random error in the estimates.
In multiplier methods, PSE = M / P
- M: the count of unique individuals from the target H2R population receiving a service or a unique object distributed among them
- P: the proportion of the target H2R population who received the service (or unique object)
The paper outlines a method for sample size calculation that incorporates the variance in M and P. They showed the high variance in P, and M impacts the precision of PSE significantly. Larger sample sizes for surveys of hard-to-reach populations are required for higher precision for PSE, especially when P is small.
Using methods described in their paper, and other resources, we created a free PSE Sample Size Calculator. Try it!
| Fearon E, Chabata S, Thompson J, Cowan F, Hargreaves J. Sample Size Calculations for Population Size Estimation Studies Using Multiplier Methods With Respondent-Driven Sampling Surveys.JMIR Public Health Surveill 2017;3(3):e59. URL: https://publichealth.jmir.org/2017/3/e59. DOI: 10.2196/publichealth.7909 | |
Methods & Quality Checklist | |
Like capture-recapture, the multiplier method employs two sources of data of which at least one needs to be representative of the target hard-to-reach (H2R) population [1]. Two variations of this method are service multiplier and unique object multiplier. In both, the total number of H2R receiving the service or unique object at a certain period, and the proportion of H2R in the subsequent survey who report using such services or receiving the unique object are used to estimate the H2R population size. |
- Can be easily integrated with surveys of H2R.
- Different multipliers can be used at the time, which yields more confidence about the plausibility of counts.
|
It has three assumptions and limitations:
- Highly dependent on the availability and quality of data collected for other purposes.
- Visiting a service may not be remembered by participants.
- Highly variable results
| If the two data sources are correlated, the counts will be underestimated and vice versa. | Multiplier Method Quality Checklist |
- The total number of H2R (unique persons) who received the service or unique objects at a certain period is accurate.
- The eligibility/definition of the H2R who received the service or unique objective at a certain period matches with the eligibility/definition of the H2R in the survey.
- The proportion of H2R in the survey who report using such services or receiving the unique object during a certain period is accurate.
- The survey of the H2R includes a sufficient representative (i.e., selected by random) number of H2R members.
- The analysis of the H2R survey (to estimate the above proportion) is weighted (if survey respondents have different probabilities of selection, which makes the sample not represent the desired H2R).
- The period of data collection for both sources is the same (i.e., if the question is about service uptake within the last three months, the same three months- the period before the interview is used for program data).
| |
[1] Johnston LG, Prybylski D, Raymond HF, Mirzazadeh A, Manopaiboon C, & McFarland W (2013). "Incorporating the service multiplier method in respondent-driven sampling surveys to estimate the size of hidden and hard-to-reach populations: Case studies from around the world." Sexually Transmitted Diseases, 40(4), p. 304-310. | |
If you are not counted, you don’t count: Estimating the number of people experiencing homelessness in Los Angeles
Principal Investigator: Paul Wesson
- $50,000 grant awarded through the UCSF Society of Hellman Fellows
- The primary objective is to conduct a population size estimation study of people experiencing homelessness (PEH) in Los Angeles County, which is estimated to have the highest per capita PEH population in California.
- We will link two data sets of PEH - the Los Angeles Homelessness Management Information System and the Los Angeles sample of the California Statewide Study of People Experiencing Homelessness (CASPEH) – and estimate the population size using capture-recapture/multiple systems estimation.
- Using a novel Targeted Minimum Loss-based Estimation (TMLE) capture-recapture estimator, we will:
- Estimate the total number of people experiencing homelessness in Los Angeles County and
- Estimate the number of people experiencing homelessness in Los Angeles County within stratified population subgroups (e.g., race and ethnicity, sex and gender, age bands).
Surveillance of injection and non-injection drug use
Principal Investigator: Paul Wesson
- This work is in response to the rise in overdose deaths in San Francisco.
- In partnership with the San Francisco Department of Public Health-Behavioral Health Services, Dr. Wesson will implement a capture-recapture/multiple systems estimation study to estimate the number of people at risk for a drug overdose in San Francisco
- The study will leverage at least four lists derived from electronic medical records, fatal drug overdoses, and non-fatal drug overdoses.
- We will use the Targeted Minimum Loss-based Estimation (TMLE) capture-recapture estimator to estimate the population size for each individual year spanning 2018 through 2022.
| Post-doctoral Fellow Opportunity | Innovations for Youth (i4Y) |
Innovations for Youth (i4Y), a youth equity research lab at UC Berkeley, has an opening for a
twelve-month, in-residence postdoctoral scholar in social epidemiology and biostatistics.
The postdoctoral scholar will collaborate with faculty at UC Berkeley and UCSF to conduct population size estimation of youth experiencing homelessness in 6 California communities: Long Beach, San Diego, San Benito/Monterey/Salinas, Oakland/Alameda County, Visalia/Kings/Tulare, and Lake County.
We are ideally seeking candidates who are available to start asap or as of July 1st, but this is
not required.
| Data Corner & Consultation |
If you are a student, post-doc, researcher, faculty, physician, or health officer, it is never too late to start your H2R research or conduct secondary data analysis and write a paper for a health topic on H2R populations.
Submit your data request online to be involved in one of our past or current projects on an H2R population.
|
We have office hours for those who want to meet virtually or in person and consult with one of our team members about an H2R population study or method.
- Tuesdays (17:00 to 18:00 PT)
- Wednesdays (16:00 to 17:00 PT)
| |
We would love to highlight any work from you, our community.
Please send us any content or suggestions you may have for future issues at H2R@ucsf.edu.
| |
Issue No. 4
This newsletter is a collaboration between staff and faculty at the University of California, San Francisco, who are focused on supporting research for hard-to-reach populations. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH, NIMH, or other funding institutes.
Issue Author: Ali Mirzazadeh, Shreya Arunsaravanakumar.
Issue Editors: Ali Mirzazadeh, Maeve Forster, Shreya Arunsaravanakumar.
Sponsor: NIH - Office of the Director
| |
Subscribe!
Please share our subscription link with your community
| | |
Keep in touch!
email: H2R@ucsf.edu
| | | | |