Validation of mobile phone use recall in the multinational MOBI-kids study

Abstract

Validation of mobile phone use recall in the multinational MOBI-kids study van Wel L, Huss A, Kromhout H, Momoli F, Krewski D, Langer CE, Castaño-Vinyals G, Kundi M, Maule M, Miligi L, Sadetzki S, Albert A, Alguacil J, Aragones N, Badia F, Bruchim R, Goedhart G, de Llobet P, Kiyohara K, Kojimahara N, Lacour B, Morales-Suarez-Varela M, Radon K, Remen T, Weinmann T, Vrijheid M, Cardis E, Vermeulen R; MOBI-Kids consortium. Validation of mobile phone use recall in the multinational MOBI-kids study. Bioelectromagnetics. 2024 May 22. doi: 10.1002/bem.22507. Abstract Potential differential and non-differential recall error in mobile phone use (MPU) in the multinational MOBI-Kids case-control study were evaluated. We compared self-reported MPU with network operator billing record data up to 3 months, 1 year, and 2 years before the interview date from 702 subjects aged between 10 and 24 years in eight countries. Spearman rank correlations, Kappa coefficients and geometric mean ratios (GMRs) were used. No material differences in MPU recall estimates between cases and controls were observed. The Spearman rank correlation coefficients between self-reported and recorded MPU in the most recent 3 months were 0.57 and 0.59 for call number and for call duration, respectively. The number of calls was on average underestimated by the participants (GMR = 0.69), while the duration of calls was overestimated (GMR = 1.59). Country, years since start of using a mobile phone, age at time of interview, and sex did not appear to influence recall accuracy for either call number or call duration. A trend in recall error was seen with level of self-reported MPU, with underestimation of use at lower levels and overestimation of use at higher levels for both number and duration of calls. Although both systematic and random errors in self-reported MPU among participants were observed, there was no evidence of differential recall error between cases and controls. Nonetheless, these sources of exposure measurement error warrant consideration in interpretation of the MOBI-Kids case-control study results on the association between children's use of mobile phones and potential brain cancer risk. Highlights • Self-reported and operator-registered phone calls were compared among MobiKids participants. • On average, number of calls was underestimated, and duration overestimated. • No differential recall error was found between brain tumor cases and controls in MobiKids. Excerpts The comparison of self-reported and recorded MPU was conducted separately for the number and duration (in minutes) of calls in the 3 months preceding the date of interview. In addition, a subset of subjects who had data available for 1 year, and a subset for 2 years preceding the date of interview were assessed.... A total of 702 subjects from eight different countries (Canada, France, Germany, Greece, Israel, Italy, Korea, and Spain) had data on both self-reported and recorded MPU in the 3 months preceding the interview date (Table 1).... The overall GMRs of self-reported versus recorded MPU in the most recent 3 months were 1.59 for call duration and 0.69 for call number, indicating systematic errors in the form of overreporting of call duration of calls and underreporting of the number of calls. Looking at recall over time for subjects with 2 years of data available, there appears to be a lower level of overreporting of duration of calls for both cases (initial GMR 1.62, 1.44 at 1 year) and controls (initial GMR 1.37, 1.08 at 1 year) between recent and 1 year recall. There is however little difference in recall in call duration over time between the 1 and 2 year time points for either cases or controls (1.44 vs. 1.41 and 1.08 vs. 1.07, respectively) (Table 5). Although this same initial decrease can be seen in controls for recall of number of calls, it is less clear for cases (Table 4).... While not all subjects from the main MOBI-Kids study provided informed consent to obtain their network operator data and not all operators provided data, we managed to include a large proportion of subjects (24.8%) from the MOBI-Kids case–control study in this validation study. The proportion of subjects where longer-term data was available (at 1- and 2-year time periods) was smaller, with no subjects from some of the participating countries.... ... it may not always be clear from billing records who the actual user of the phone was when making calls. Therefore, while billing records provide insight into validity and possible calibration of self-reports, they may not represent the gold standard for studies, especially due to the low number of retained study participants, and also because internet-based calling (VoiP) is not included in the records.... Both the COSMOS (Reedijk et al., 2024) as well as the INTERPHONE validation (Vrijheid et al., 2009) studies found significantly differing ratios between countries. In the present study we did see some differences in GMRs among the eight participating countries, but the differences did not achieve statistical significance. In contrast, the MOBI-Expo validation study by Goedhart et al. (2018) did find significant differences among countries with participants from Greece, Israel, and Korea underestimating the duration of calls while in other countries duration was overestimated.... CONCLUSIONS We compared self-reported MPU with operator data at 3 months, 1 year, and 2 years preceding the interview date. No indication of differential recall error between cases and controls was found. Both non-differential systematic and random errors were observed, with number of calls being underreported and duration of calls being over-reported on average in both cases and controls. If there are true underlying risks, then the observed non-differential random errors may bias risk estimates towards their null values and decrease study power. Open access paper: onlinelibrary.wiley.com

AI evidence extraction

Main findings

Among 702 participants aged 10–24 years in eight countries, self-reported mobile phone use was compared with operator billing records. Spearman correlations for the most recent 3 months were 0.57 (call number) and 0.59 (call duration). On average, participants underestimated number of calls (GMR 0.69) and overestimated call duration (GMR 1.59), with no material differences in recall estimates between cases and controls (no evidence of differential recall error).

Outcomes measured

• Validity/accuracy of self-reported mobile phone use (call number, call duration) versus network operator billing records
• Differential vs non-differential recall error between cases and controls

Limitations

• Not all MOBI-Kids subjects consented to operator data and not all operators provided data; validation sample represented 24.8% of the main study
• Smaller subsets had 1-year and 2-year operator data; some countries had no longer-term data
• Billing records may not reflect the actual phone user for calls
• Internet-based calling (VoIP) not included in operator billing records
• Systematic and random errors in self-reported mobile phone use were observed

Suggested hubs

• mobile-phones (0.9)
  Validates self-reported mobile phone use against operator records in MOBI-Kids participants.
• exposure-assessment (0.85)
  Focus is on exposure measurement error/recall accuracy for mobile phone use.

{
    "study_type": "exposure_assessment",
    "exposure": {
        "band": null,
        "source": "mobile phone",
        "frequency_mhz": null,
        "sar_wkg": null,
        "duration": "Self-reported mobile phone use compared with operator billing records for periods up to 3 months, 1 year, and 2 years before interview"
    },
    "population": "Subjects aged 10–24 years from the multinational MOBI-Kids case-control study (eight countries)",
    "sample_size": 702,
    "outcomes": [
        "Validity/accuracy of self-reported mobile phone use (call number, call duration) versus network operator billing records",
        "Differential vs non-differential recall error between cases and controls"
    ],
    "main_findings": "Among 702 participants aged 10–24 years in eight countries, self-reported mobile phone use was compared with operator billing records. Spearman correlations for the most recent 3 months were 0.57 (call number) and 0.59 (call duration). On average, participants underestimated number of calls (GMR 0.69) and overestimated call duration (GMR 1.59), with no material differences in recall estimates between cases and controls (no evidence of differential recall error).",
    "effect_direction": "no_effect",
    "limitations": [
        "Not all MOBI-Kids subjects consented to operator data and not all operators provided data; validation sample represented 24.8% of the main study",
        "Smaller subsets had 1-year and 2-year operator data; some countries had no longer-term data",
        "Billing records may not reflect the actual phone user for calls",
        "Internet-based calling (VoIP) not included in operator billing records",
        "Systematic and random errors in self-reported mobile phone use were observed"
    ],
    "evidence_strength": "moderate",
    "confidence": 0.7800000000000000266453525910037569701671600341796875,
    "peer_reviewed_likely": "yes",
    "keywords": [
        "MOBI-Kids",
        "mobile phone use",
        "recall validation",
        "operator billing records",
        "exposure measurement error",
        "case-control",
        "children and young adults",
        "brain tumor cases and controls"
    ],
    "suggested_hubs": [
        {
            "slug": "mobile-phones",
            "weight": 0.90000000000000002220446049250313080847263336181640625,
            "reason": "Validates self-reported mobile phone use against operator records in MOBI-Kids participants."
        },
        {
            "slug": "exposure-assessment",
            "weight": 0.84999999999999997779553950749686919152736663818359375,
            "reason": "Focus is on exposure measurement error/recall accuracy for mobile phone use."
        }
    ]
}

Comments

Log in to comment.

No comments yet.